Measurement trends in psychology

Analysis

TODO

• update prior in this file and the other to match prior justification
• prevalence-and-trend plot might need to choose between marginal and conditional means. perhaps component plots too.
• try to reproduce “” warning/error in prevalence-and-trend plots and fix
• Add an (alternative) weighted model that weights by total articles published in each journal during that year? This should only produce results for the field as a whole, eg “Psychology (weighted by publications per journal per year)”
• remove/refactor plotting of summary data in plots, as the real data will be planned-missing

library(dplyr)
library(tidyr)
library(tibble)
library(stringr)
library(forcats)
library(readxl)
library(purrr)
library(readr)
library(janitor)
library(ggplot2)
library(scales)
library(ggstance)
library(ggcorrplot)
library(brms)
library(tidybayes)
library(emmeans)
library(marginaleffects)
library(knitr)
library(kableExtra)


dir.create("models")
dir.create("plots")

# function to save ggplots to disk
quicksave <- function(name, width, height){
  ggsave(paste0("plots/", name, ".pdf"), 
         device = "pdf",
         width = width, 
         height = height)
  
  ggsave(paste0("plots/", name, ".png"), 
         device = "png",
         dpi = 600,
         width = width, 
         height = height)
}

# round using half-up strategy and retaining trailing zeros
rnd <- function(x, digits = 2) {
  rounded_number <- round_half_up(x, digits)
  formatted_number <- sprintf(paste0("%.", digits, "f"), rounded_number)
  return(formatted_number)
}

Data

Simulate data

We will have two raters and they can either (a) score each article so that each one is scored twice, or (b) they could overlap eg 10% of articles for double scoring, so that more articles can be scored in total. Any input on this?

Each rater will score each measurement type as either present (1) or absent (0) in that article. There are 6 measurement types according to our scoring, and we see these categories as exhaustive. There will probably be 6 subfields and 5 journals per subfield.

In the case of disagreements, should we solve this in data processing (get agreement between the authors/round down/round up) or model it?

The total number of articles to be scored is still to be determined, as we don’t know the average extraction time or the total resources available to us here (1 or 2 RAs or masters students or a mix).

When a journal has not existed for all years in the range, should we still extract the same number of articles from the years that do exist, or should we extract a proportionate number of articles (so the articles per year is stable)?

set.seed(42)

n_articles_per_journal <- 15 # current estimate
n_journals_per_subfield <- 5
n_subfields <- 6

n_articles_total <- n_articles_per_journal * n_journals_per_subfield * n_subfields

data_simulated <-
  tibble(
    article_id = seq(from = 1, 
                     to = n_articles_total, 
                     by = 1),
    year = sample(x = 2009:2023,  # year range 
                  size = n_articles_total, 
                  replace = TRUE),
    directbehavioral = rbinom(n = n_articles_total,
                              size = 1, # values between 0 and 1
                              prob = .05), # probability of a given study containing such a measure
    behavioralproxy = rbinom(n = n_articles_total,
                             size = 1, # values between 0 and 1
                             prob = .1), # probability of a given study containing such a measure
    selfreportsaboutbehavior = rbinom(n = n_articles_total,
                                      size = 1, # values between 0 and 1
                                      prob = .1), # probability of a given study containing such a measure
    selfreportaboutsubjectivestate = c(rbinom(n = n_articles_total/2,
                          size = 1, # values between 0 and 1
                          prob = .95), # probability of a given study containing such a measure
                   rbinom(n = n_articles_total/2,
                          size = 1, # values between 0 and 1
                          prob = .70)),
    neurobiopsychophys = rbinom(n = n_articles_total,
                       size = 1, # values between 0 and 1
                       prob = .1), # probability of a given study containing such a measure
    cognitiveability = rbinom(n = n_articles_total,
                              size = 1, # values between 0 and 1
                              prob = .1) # probability of a given study containing such a measure
  ) |>
  mutate(article_id = as.factor(paste0("doi_", article_id)),
         subfield = c(rep("General", n_articles_per_journal * n_journals_per_subfield),
                      rep("Cognitive", n_articles_per_journal * n_journals_per_subfield),
                      rep("Clinical", n_articles_per_journal * n_journals_per_subfield),
                      rep("Developmental", n_articles_per_journal * n_journals_per_subfield),
                      rep("Industrial & Organizational", n_articles_per_journal * n_journals_per_subfield),
                      rep("Social & Personality", n_articles_per_journal * n_journals_per_subfield)),
         subfield = as.factor(subfield),
         journal_label = paste(subfield, "journal", seq(from = 1, to = 5)))

# weights
data_weights <- data_simulated |>
  distinct(subfield, journal_label) |>
  mutate(total_articles = round(runif(n = n(), min = 10*4, max = 10*12), 0),
         total_articles = case_when(subfield == "General" ~ total_articles + 80,
                                    subfield == "Clinical" ~ total_articles + 40,
                                    TRUE ~ total_articles),
         percent_of_all_articles = total_articles/sum(total_articles)*100)

data_simulated <- data_simulated |>
  left_join(data_weights, by = c("subfield", "journal_label"))

journal_names <- read_xlsx("../../data/registered report/journals.xlsx")

data_simulated <- data_simulated |>
  left_join(journal_names, by = c("subfield", "journal_label")) |>
  rename(journal = journal_abbreviation)

# save for reproducibility
if(!file.exists("../../data/registered report/simulated/data_simulated.csv")){
  write_csv(data_simulated, "../../data/registered report/simulated/data_simulated.csv")
}

# pull the journal abbreviations, for ordering figures and tables later
subfield_and_journal_labels <- journal_names |>
  pull(journal_abbreviation)

subfield_and_journal_labels_without_subfields <- journal_names |>
  drop_na(journal_name) |>
  pull(journal_abbreviation)

Check

data_simulated |>
  count(subfield, journal) |>
  kable() |>
  kable_classic(full_width = FALSE)

subfield	journal	n
Clinical	BRAT	15
Clinical	CPS	15
Clinical	JCCAP	15
Clinical	JCCP	15
Clinical	JPCS	15
Cognitive	Cognition	15
Cognitive	JEP:HPP	15
Cognitive	JEP:LMC	15
Cognitive	JML	15
Cognitive	M&C	15
Developmental	CD	15
Developmental	DP	15
Developmental	JECP	15
Developmental	JRA	15
Developmental	PA	15
General	Collabra	15
General	JEP:G	15
General	NHB	15
General	PLOSONE	15
General	PS	15
Industrial & Organizational	JAP	15
Industrial & Organizational	JEP:A	15
Industrial & Organizational	JOB	15
Industrial & Organizational	OBHDP	15
Industrial & Organizational	PP	15
Social & Personality	Emotion	15
Social & Personality	JESP	15
Social & Personality	JPSP	15
Social & Personality	JRP	15
Social & Personality	PSPB	15

data_simulated |>
  summarize(mean_directbehavioral = mean(directbehavioral),
            mean_behavioralproxy = mean(behavioralproxy),
            mean_selfreportsaboutbehavior = mean(selfreportsaboutbehavior),
            mean_selfreportaboutsubjectivestate = mean(selfreportaboutsubjectivestate),
            mean_neurobiopsychophys = mean(neurobiopsychophys),
            mean_cognitiveability = mean(cognitiveability)) |>
  mutate_all(round_half_up, digits = 3) |>
  kable() |>
  kable_classic(full_width = FALSE)

mean_directbehavioral	mean_behavioralproxy	mean_selfreportsaboutbehavior	mean_selfreportaboutsubjectivestate	mean_neurobiopsychophys	mean_cognitiveability
0.04	0.089	0.107	0.833	0.122	0.078

data_simulated |>
  group_by(subfield, journal) |>
  summarize(mean_directbehavioral = mean(directbehavioral),
            mean_behavioralproxy = mean(behavioralproxy),
            mean_selfreportsaboutbehavior = mean(selfreportsaboutbehavior),
            mean_selfreportaboutsubjectivestate = mean(selfreportaboutsubjectivestate),
            mean_neurobiopsychophys = mean(neurobiopsychophys),
            mean_cognitiveability = mean(cognitiveability)) |>
  mutate_if(is.numeric, round_half_up, digits = 3) |>
  kable() |>
  kable_classic(full_width = FALSE)

subfield	journal	mean_directbehavioral	mean_behavioralproxy	mean_selfreportsaboutbehavior	mean_selfreportaboutsubjectivestate	mean_neurobiopsychophys	mean_cognitiveability
Clinical	BRAT	0.067	0.333	0.133	0.933	0.200	0.000
Clinical	CPS	0.000	0.067	0.067	1.000	0.200	0.133
Clinical	JCCAP	0.133	0.133	0.067	1.000	0.067	0.067
Clinical	JCCP	0.000	0.067	0.000	0.733	0.067	0.067
Clinical	JPCS	0.000	0.067	0.133	0.933	0.067	0.133
Cognitive	Cognition	0.067	0.067	0.133	0.933	0.000	0.067
Cognitive	JEP:HPP	0.133	0.000	0.133	0.867	0.067	0.067
Cognitive	JEP:LMC	0.067	0.133	0.067	0.933	0.067	0.000
Cognitive	JML	0.000	0.067	0.133	0.933	0.133	0.133
Cognitive	M&C	0.000	0.133	0.133	0.933	0.267	0.067
Developmental	CD	0.067	0.067	0.200	0.800	0.067	0.067
Developmental	DP	0.067	0.133	0.200	0.800	0.133	0.000
Developmental	JECP	0.067	0.000	0.200	0.733	0.000	0.067
Developmental	JRA	0.000	0.200	0.200	0.800	0.200	0.133
Developmental	PA	0.067	0.200	0.000	0.667	0.067	0.067
General	Collabra	0.067	0.133	0.067	1.000	0.267	0.067
General	JEP:G	0.000	0.133	0.267	1.000	0.200	0.067
General	NHB	0.067	0.067	0.000	0.867	0.133	0.067
General	PLOSONE	0.000	0.000	0.067	0.933	0.200	0.133
General	PS	0.067	0.067	0.067	1.000	0.067	0.000
Industrial & Organizational	JAP	0.000	0.000	0.000	0.800	0.067	0.200
Industrial & Organizational	JEP:A	0.000	0.000	0.067	0.733	0.200	0.133
Industrial & Organizational	JOB	0.000	0.133	0.067	0.600	0.133	0.000
Industrial & Organizational	OBHDP	0.067	0.133	0.067	0.867	0.067	0.067
Industrial & Organizational	PP	0.067	0.000	0.000	0.533	0.200	0.067
Social & Personality	Emotion	0.067	0.067	0.067	0.533	0.000	0.267
Social & Personality	JESP	0.000	0.000	0.133	0.800	0.000	0.000
Social & Personality	JPSP	0.000	0.267	0.267	0.667	0.067	0.000
Social & Personality	JRP	0.000	0.000	0.133	0.733	0.267	0.067
Social & Personality	PSPB	0.067	0.000	0.133	0.933	0.200	0.133

Real data

# simulated data will be replaced with real data after collection
# data_processed <- read_csv("../../data/processed/data_processed.csv")

Analyze

Justification of priors

Intuitions for log-odds

We employ a logistic regression (Bernoulli link function), so priors for the intercepts, b_year_centered, and sd of the random effects are on the log-odds scale.

To illustrate and develop an intuition for log-odds values, we can convert probabilities to log-odds and table/plot them against one another.

df_probability_logodds <- 
  tibble(probability = seq(0, 1, by = 0.001)) |>
  mutate(logodds = qlogis(probability))

df_probability_logodds |>
  filter(probability %in% c(0.01, 0.5, 0.99)) |>
  kable() |>
  kable_classic(full_width = FALSE)

probability	logodds
0.01	-4.59512
0.50	0.00000
0.99	4.59512

ggplot(df_probability_logodds, aes(probability, logodds)) +
  geom_point(size = 0.1) +
  geom_vline(xintercept = 0.01, linetype = "dashed", color = "#359EAAFF") +
  geom_vline(xintercept = 0.99, linetype = "dashed", color = "#359EAAFF") +
  theme_linedraw() +
  scale_x_continuous(breaks = breaks_pretty(n = 10)) +
  xlab("Probability") +
  ylab("Log-odds")

This shows that a probability of 50% = log-odds of 0. More importantly, |log-odds| > 4.6 refer to probabilities of less than 1%. So, to say that there is a log-odds ±5, 6, or 7 of something occurring means that it almost never/always occurs. For our study, prevalences in the 1-99% range are reasonable.

Intercepts (prevalences)

We are looking to employ weakly regularizing priors, i.e., those that help the model converge, and therefore exclude impossible/highly implausible values, without imposing strong prior beliefs onto the model. Given this goal, our priors should therefore both cover the probability of 1-99% range (and a bit beyond) and exclude probabilities that approach ±Inf, therefore |log-odds| of ±4.6 or so. A normally distributed prior on the log-odds scale is therefore useful here.

We will choose specific locations based our a priori beliefs about the prevalence of different types of measures. However, given the width of the prior on each, this won’t make much difference. It’s more about indicating to the model what is unlikely (e.g., only 1% of studies using self-report measures) rather than a strong belief about what is likely.

Specifically, we choose three different priors depending on our beliefs about each type of measure:

• No strong beliefs about prevalence: mean = .50 (on probability scale), sd = 1.4 (on log-odds scale). This prior serves only to specify that very high or low prevalences are less likely.
• Prior belief prevalence is high: mean = .70 (on probability scale), sd = 1 (on log-odds scale). This prior serves to specify that low prevalences less likely.
• Prior belief prevalence is low: mean = .25 (on probability scale), sd = 1 (on log-odds scale). This prior serves to specify that high prevalences less likely.

The asymmetry between low and high is intentional, reflecting our stronger belief that low prevalences are lower than high prevalence are high.

illustrate_intercept_prior <- function(mean_prob, sd_logodds){
  df_normal_logodds <- 
    tibble(logodds = rnorm(mean = qlogis(mean_prob), sd = sd_logodds, n = 10000000)) |>
    mutate(probability = plogis(logodds))
  
  ggplot(df_normal_logodds, aes(probability)) +
    geom_density(fill = "grey20") +
    theme_linedraw() +
    scale_x_continuous(breaks = breaks_pretty(n = 10)) +
    xlab("Probability") +
    ylab("Density") +
    ggtitle(paste0("Prevalence prior: \nnormal(M = qlogis(", mean_prob, "), SD = 1) log-odds distribution converted to probabilities"))
}

illustrate_intercept_prior(mean_prob = .70, sd_logodds = 1) # self reports of subjective experience

illustrate_intercept_prior(mean_prob = .25, sd_logodds = 1) # direct behavioral measures

illustrate_intercept_prior(mean_prob = .50, sd_logodds = 1.4) # all other modes, given our very weak beliefs and the potential for variation between subfields/journals 

qlogis(.50)

## [1] 0

qlogis(.70)

## [1] 0.8472979

qlogis(.25)

## [1] -1.098612

Slopes (trends)

Normal distributions (on the log-odds scale) are also useful for the slopes. However, because the slopes compound over years, relatively small values of slope can produce large changes in the prevalence over the 15 year period.

The below visualizes different (a) values of prevalence (probability scale) in the middle year (the mean of the prior distribution) and (b) values of 2 X SD of the prior distribution (on the log-odds scale), which corresponds to roughly 95% of the range of values covered by the prior.

illustrate_trend_prior <- function(mean_prevalence_prob, sd_trend_prob){
  df_probability_logodds <- 
    tibble(year_centered = seq(-7, +7, by = 1)) |>
    expand_grid(b1 = seq(sd_trend_prob*-2, sd_trend_prob*2, by = .01)) |>
    mutate(logodds = qlogis(mean_prevalence_prob) + (b1 * year_centered),
           probability = plogis(logodds),
           year = year_centered + 2016)
  
  ggplot(df_probability_logodds, aes(year, probability, color = b1, group = b1)) +
    geom_line() +
    scale_x_continuous(breaks = breaks_pretty(n = 10)) +
    theme_linedraw() +
    ylim(0, 1) +
    ylab("Prevalence") +
    xlab("Year") +
    ggtitle(paste0("Prevalence prior: \nnormal(M = qlogis(", mean_prevalence_prob, "), SD = ", sd_trend_prob, ") log-odds distribution converted to probabilities"))
}

illustrate_trend_prior(mean_prevalence_prob = .10, sd_trend_prob = .1)

illustrate_trend_prior(mean_prevalence_prob = .50, sd_trend_prob = .1)

illustrate_trend_prior(mean_prevalence_prob = .90, sd_trend_prob = .1)

illustrate_trend_prior(mean_prevalence_prob = .10, sd_trend_prob = .2)

illustrate_trend_prior(mean_prevalence_prob = .50, sd_trend_prob = .2) # probably for the best

illustrate_trend_prior(mean_prevalence_prob = .90, sd_trend_prob = .2)

illustrate_trend_prior(mean_prevalence_prob = .10, sd_trend_prob = .3)

illustrate_trend_prior(mean_prevalence_prob = .50, sd_trend_prob = .3)

illustrate_trend_prior(mean_prevalence_prob = .90, sd_trend_prob = .3)

A normal prior for the slopes with M = 0 and SD = .20 allows, regardless of the prevalence in the middle year, for very large trends across years. E.g., for prevalence = .50 in the middle year, the 95% of the prior allows for changes of from roughly 5% in the first year to 95% in the final year. This is unlikely, but the point of the these weakly regularising priors is to aid convergence by excluding truly impossible or highly implausible values. This prior does this.

SD of the random effects

The width of the SDs on the random effects (subfield and subfield:journal) also require a prior, still on the log-odds scale, which refers to the degree of variability between subfields/journals.

McElreath uses exponential(1) as a weakly regularizing prior for SDs on random effects frequently in the 2nd edition of Statistical Rethinking, and therefore we do here. Again, this serves to exclude truly impossible or highly implausible values (e.g., 10 or 100) to aid convergence, rather than to assert strong prior beliefs. The below plots such an exp(1) distribution.

prior_sd <- tibble(sd = seq(0, 4, by = 0.1)) |>
  mutate(density = dexp(sd, rate = 1))  

ggplot(prior_sd, aes(sd, density)) +
  geom_area() +
  xlab("SD of the random effects (on the log-odds scale)") +
  ylab("Density") +
  theme_linedraw()

Correlations among the random effects

{brms} uses a Lewandowski-Kurowicka-Joe (LKJ) prior with eta = 1, i.e., a flat distribution. LKJ depends on the number of correlations involved, and we have a relatively large amount (21), so values of eta above 1, which would regularize correlations towards zero, would be more aggressive than the weakly regularising prior that we’re looking for. LKJ(1) is merely skeptical of large correlations, but is applied to the matrix as a whole, so a minority of large correlations are still possible with this prior. We therefore retain the {brms} default prior of lkj(1).

Fit model

# # real data
# data_centered <- data_processed |>
#   mutate(year_centered = year - 2016)

# simulated data for development
data_centered <- data_simulated |>
  mutate(year_centered = year - 2016)

# specify Wilkinson notation for the model
# note that the | p | and | q | syntax specifies that the correlations among the random effects be modelled (ie not assumed to be zero)
formulas <- 
  bf(formula = directbehavioral | weights(percent_of_all_articles) ~               1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
  bf(formula = behavioralproxy | weights(percent_of_all_articles) ~                1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
  bf(formula = selfreportsaboutbehavior | weights(percent_of_all_articles) ~       1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
  bf(formula = selfreportaboutsubjectivestate | weights(percent_of_all_articles) ~ 1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
  bf(formula = neurobiopsychophys | weights(percent_of_all_articles) ~             1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
  bf(formula = cognitiveability | weights(percent_of_all_articles) ~               1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal))

# # it is plausible that a reviewer asks us for unweighted models for robustness, although it is harder to know how to interpret the field-wide prevalences and trends given meaningful differences in size of the subfields. nonetheless, I retain the code here.
# formulas <- 
#   bf(formula = directbehavioral ~               1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
#   bf(formula = behavioralproxy ~                1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
#   bf(formula = selfreportsaboutbehavior ~       1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
#   bf(formula = selfreportaboutsubjectivestate ~ 1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
#   bf(formula = neurobiopsychophys ~             1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) +
#   bf(formula = cognitiveability ~               1 + year_centered + (1 + year_centered | p | subfield) + (1 + year_centered | q | subfield:journal)) 

# specify priors
priors <- 
  prior(normal(-1.10, 1.0), class = Intercept, resp = "directbehavioral") + # qlogis(.25)
  prior(normal(+0.00, 1.4), class = Intercept, resp = "behavioralproxy") + # qlogis(.50)
  prior(normal(+0.00, 1.4), class = Intercept, resp = "selfreportsaboutbehavior") + # qlogis(.50)
  prior(normal(+0.85, 1.0), class = Intercept, resp = "selfreportaboutsubjectivestate") + # qlogis(.70)
  prior(normal(+0.00, 1.4), class = Intercept, resp = "neurobiopsychophys") + # qlogis(.50)
  prior(normal(+0.00, 1.4), class = Intercept, resp = "cognitiveability") + # qlogis(.50)
  
  prior(normal(0, 0.2),   class = b,         resp = "directbehavioral") +
  prior(normal(0, 0.2),   class = b,         resp = "behavioralproxy") + 
  prior(normal(0, 0.2),   class = b,         resp = "selfreportsaboutbehavior") + 
  prior(normal(0, 0.2),   class = b,         resp = "selfreportaboutsubjectivestate") + 
  prior(normal(0, 0.2),   class = b,         resp = "neurobiopsychophys") + 
  prior(normal(0, 0.2),   class = b,         resp = "cognitiveability") + 
  
  prior(exponential(1),   class = sd,        resp = "directbehavioral") +
  prior(exponential(1),   class = sd,        resp = "behavioralproxy") + 
  prior(exponential(1),   class = sd,        resp = "selfreportsaboutbehavior") + 
  prior(exponential(1),   class = sd,        resp = "selfreportaboutsubjectivestate") + 
  prior(exponential(1),   class = sd,        resp = "neurobiopsychophys") + 
  prior(exponential(1),   class = sd,        resp = "cognitiveability") + 
  
  prior(lkj_corr_cholesky(1), class = cor)

# fit model
fitted_model <- brm(
  formula = formulas,
  family  = bernoulli(link = "logit"),
  prior   = priors,
  data    = data_centered,
  cores   = parallel::detectCores(),
  seed    = 42,
  iter    = 8000, 
  warmup  = 4000, 
  
  # sample prior only for testing and dev
  #sample_prior = "only",
  #file = "models/fitted_model_prioronly"
  
  # fit using simulated data for testing and dev
  sample_prior = "no",
  file = "models/fitted_model_simulated_data"
  
  # when real data is in after Stage 1 acceptance, fit using real data
  #sample_prior = "no",
  #file = "models/fitted_model"
  
  # if there are divergent transitions, can be increased from the default to eg .99 
  #control = list(adapt_delta = .99)
)

Model diagnostics

Extract estimates

Resources

For consistent and precise terminology, we relied on Heiss (2021) and Heiss (2022) which describe marginal vs conditional effects implemented in the marginaleffects vs emmeans packages.

Following Heiss’s (2021, 2022) terminology, we extract:

• Marginal Global Mean (aka the Average Marginal Effect, AME): “typically, psychology articles…” - intervals are akin to confidence intervals in Frequentist meta analysis
• Conditional Global Mean (aka the Group Average Marginal Effect, G-AME)): “a typical psychology article…” - intervals are akin to prediction intervals in Frequentist meta analysis, will be equal or wider than the AME estimates.
• Conditional Mean by Subfield
• Conditional Mean by Journal

Prevalence

if(exists("models/posterior_prevalence.rds")){
  posterior_prevalence <- read_rds("models/posterior_prevalence.rds")
} else {
  
  est_prevalence <- function(subfield = "hypothetical typical subfield", journal = "hypothetical typical journal", type = c("conditional", "marginal"), year_centered = 0, fit) {
    
    # reminder of the distinction: 
    # "a typical journal..." (conditional effect, taking levels of the random effects into account, whether existing or new hypothetical ones) vs.
    # "typically, journals..." (marginal effect, ignoring the random effects, therefore likely producing wider estimates given any heterogeneity among the random effects)
    
    if(type == "conditional") {
      # conditional effect for either existing levels of the random effects if specified,
      # or if not specified then for hypothetical new representative levels 
      if(subfield == "hypothetical typical subfield"){
        subfield_dummy = factor("hypothetical typical subfield", levels = unique(fit$data$subfield))
      } else {
        subfield_dummy <- subfield
      }
      
      if(journal == "hypothetical typical journal"){
        journal_dummy = factor("hypothetical typical journal", levels = unique(fit$data$journal))
      } else {
        journal_dummy <- journal
      }
      
      res <- 
        predictions(model = fitted_model,
                    newdata = datagrid(year_centered = year_centered,
                                       subfield = subfield_dummy,
                                       journal = journal_dummy),
                    re_formula = NULL,
                    allow_new_levels = TRUE) |>
        select(outcome = group, 
               prevalence_estimate = estimate,
               prevalence_ci_lower = conf.low,
               prevalence_ci_upper = conf.high)
      
    } else if (type == "marginal") {
      res <- 
        predictions(model = fit,
                    newdata = datagrid(year_centered = year_centered),
                    re_formula = NA,  
                    allow_new_levels = FALSE) |>
        select(outcome = group, 
               prevalence_estimate = estimate,
               prevalence_ci_lower = conf.low,
               prevalence_ci_upper = conf.high)
    }
    
    return(res)
  }
  
  # field - marginal effect
  posterior_prevalence_field_marginal <- 
    expand_grid(subfield = "across field", 
                journal = "across field",
                type = "marginal",
                re_level = "field") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # field - conditional effect
  posterior_prevalence_field_conditional <- 
    expand_grid(subfield = "hypothetical typical subfield", 
                journal = "hypothetical typical journal",
                type = "conditional",
                re_level = "field") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # subfields
  posterior_prevalence_subfields <- 
    expand_grid(data_centered |>
                  distinct(subfield),
                journal = "hypothetical typical journal",
                type = "conditional",
                re_level = "subfields") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # journals
  posterior_prevalence_journals <- 
    expand_grid(data_centered |>
                  distinct(subfield, journal),
                type = "conditional",
                re_level = "journals") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # combine
  posterior_prevalence <- 
    bind_rows(posterior_prevalence_field_marginal,
              posterior_prevalence_field_conditional,
              posterior_prevalence_subfields,
              posterior_prevalence_journals) |>
    mutate(label = case_when(re_level == "field" & type == "marginal" ~ "Psychology (marginal)",
                             re_level == "field" & type == "conditional" ~ "Psychology (conditional)",
                             re_level == "subfields" & journal == "hypothetical typical journal" ~ subfield,
                             journal != "hypothetical typical journal" & !is.na(journal) ~ journal)) |>
    mutate(label = fct_relevel(label,
                               "Psychology (marginal)",
                               "Psychology (conditional)",
                               subfield_and_journal_labels),
    label = fct_rev(label)) |>
    mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",     
                               outcome == "behavioralproxy" ~ "Behavioral Proxy",
                               outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                               outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                               outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                               outcome == "cognitiveability" ~ "Cognitive Ability",
                               TRUE ~ outcome),
           outcome = fct_relevel(outcome, 
                                 "Direct Behavioral",       
                                 "Behavioral Proxy",
                                 "Self-Report about Behavior",
                                 "Self-Report about Subjective States",
                                 "Neuro/Bio/Psychophys",
                                 "Cognitive Ability"),
           outcome = fct_rev(outcome),
           subfield = fct_relevel(subfield,
                                  "hypothetical typical subfield",
                                  "General",
                                  "Clinical",
                                  "Cognitive",
                                  "Developmental",
                                  "Industrial & Organizational",
                                  "Social & Personality")) |>
    mutate(prevalence_category = case_when(prevalence_estimate < .05 ~ "Rare",
                                           prevalence_estimate >= .05 & prevalence_estimate < .10 ~ "Uncommon",
                                           prevalence_estimate >= .10 & prevalence_estimate < .25 ~ "Occasional",
                                           prevalence_estimate >= .25 & prevalence_estimate < .50 ~ "Common",
                                           prevalence_estimate >= .50 ~ "Frequent"),
           prevalence_results = paste0(rnd(prevalence_estimate), ", 95% CI [", rnd(prevalence_ci_lower), ", ", rnd(prevalence_ci_upper), "]")) |>
    select(outcome, label, re_level, type, subfield, journal, 
           prevalence_estimate, prevalence_ci_lower, prevalence_ci_upper, prevalence_category, prevalence_results) |>
    arrange(outcome, desc(label), subfield)
  
  write_rds(posterior_prevalence, "models/posterior_prevalence.rds")
}

# print table
posterior_prevalence |>
  mutate_if(is.numeric, round_half_up, digits = 3) |>
  kable() |>
  kable_classic(full_width = FALSE)

outcome	label	re_level	type	subfield	journal	prevalence_estimate	prevalence_ci_lower	prevalence_ci_upper	prevalence_category	prevalence_results
Cognitive Ability	Psychology (marginal)	field	marginal	across field	across field	0.046	0.025	0.080	Rare	0.05, 95% CI [0.03, 0.08]
Cognitive Ability	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.046	0.005	0.225	Rare	0.05, 95% CI [0.00, 0.22]
Cognitive Ability	General	subfields	conditional	General	hypothetical typical journal	0.045	0.004	0.216	Rare	0.04, 95% CI [0.00, 0.22]
Cognitive Ability	Collabra	journals	conditional	General	Collabra	0.057	0.021	0.124	Uncommon	0.06, 95% CI [0.02, 0.12]
Cognitive Ability	JEP:G	journals	conditional	General	JEP:G	0.048	0.016	0.111	Rare	0.05, 95% CI [0.02, 0.11]
Cognitive Ability	NHB	journals	conditional	General	NHB	0.025	0.004	0.072	Rare	0.02, 95% CI [0.00, 0.07]
Cognitive Ability	PLOSONE	journals	conditional	General	PLOSONE	0.101	0.045	0.188	Occasional	0.10, 95% CI [0.04, 0.19]
Cognitive Ability	PS	journals	conditional	General	PS	0.013	0.002	0.044	Rare	0.01, 95% CI [0.00, 0.04]
Cognitive Ability	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.046	0.005	0.218	Rare	0.05, 95% CI [0.00, 0.22]
Cognitive Ability	BRAT	journals	conditional	Clinical	BRAT	0.011	0.002	0.041	Rare	0.01, 95% CI [0.00, 0.04]
Cognitive Ability	CPS	journals	conditional	Clinical	CPS	0.108	0.054	0.189	Occasional	0.11, 95% CI [0.05, 0.19]
Cognitive Ability	JCCAP	journals	conditional	Clinical	JCCAP	0.026	0.005	0.076	Rare	0.03, 95% CI [0.00, 0.08]
Cognitive Ability	JCCP	journals	conditional	Clinical	JCCP	0.044	0.012	0.109	Rare	0.04, 95% CI [0.01, 0.11]
Cognitive Ability	JPCS	journals	conditional	Clinical	JPCS	0.086	0.034	0.170	Uncommon	0.09, 95% CI [0.03, 0.17]
Cognitive Ability	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.046	0.005	0.219	Rare	0.05, 95% CI [0.00, 0.22]
Cognitive Ability	Cognition	journals	conditional	Cognitive	Cognition	0.037	0.009	0.100	Rare	0.04, 95% CI [0.01, 0.10]
Cognitive Ability	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.053	0.015	0.142	Uncommon	0.05, 95% CI [0.01, 0.14]
Cognitive Ability	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.018	0.002	0.060	Rare	0.02, 95% CI [0.00, 0.06]
Cognitive Ability	JML	journals	conditional	Cognitive	JML	0.079	0.024	0.201	Uncommon	0.08, 95% CI [0.02, 0.20]
Cognitive Ability	M&C	journals	conditional	Cognitive	M&C	0.049	0.016	0.118	Rare	0.05, 95% CI [0.02, 0.12]
Cognitive Ability	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.046	0.005	0.226	Rare	0.05, 95% CI [0.00, 0.23]
Cognitive Ability	CD	journals	conditional	Developmental	CD	0.042	0.011	0.114	Rare	0.04, 95% CI [0.01, 0.11]
Cognitive Ability	DP	journals	conditional	Developmental	DP	0.021	0.003	0.074	Rare	0.02, 95% CI [0.00, 0.07]
Cognitive Ability	JECP	journals	conditional	Developmental	JECP	0.041	0.009	0.111	Rare	0.04, 95% CI [0.01, 0.11]
Cognitive Ability	JRA	journals	conditional	Developmental	JRA	0.070	0.019	0.195	Uncommon	0.07, 95% CI [0.02, 0.19]
Cognitive Ability	PA	journals	conditional	Developmental	PA	0.035	0.008	0.106	Rare	0.03, 95% CI [0.01, 0.11]
Cognitive Ability	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.048	0.005	0.231	Rare	0.05, 95% CI [0.00, 0.23]
Cognitive Ability	JAP	journals	conditional	Industrial & Organizational	JAP	0.132	0.050	0.271	Occasional	0.13, 95% CI [0.05, 0.27]
Cognitive Ability	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.103	0.035	0.228	Occasional	0.10, 95% CI [0.04, 0.23]
Cognitive Ability	JOB	journals	conditional	Industrial & Organizational	JOB	0.019	0.003	0.063	Rare	0.02, 95% CI [0.00, 0.06]
Cognitive Ability	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.033	0.006	0.113	Rare	0.03, 95% CI [0.01, 0.11]
Cognitive Ability	PP	journals	conditional	Industrial & Organizational	PP	0.063	0.017	0.170	Uncommon	0.06, 95% CI [0.02, 0.17]
Cognitive Ability	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.048	0.005	0.228	Rare	0.05, 95% CI [0.00, 0.23]
Cognitive Ability	Emotion	journals	conditional	Social & Personality	Emotion	0.219	0.123	0.341	Occasional	0.22, 95% CI [0.12, 0.34]
Cognitive Ability	JESP	journals	conditional	Social & Personality	JESP	0.030	0.004	0.102	Rare	0.03, 95% CI [0.00, 0.10]
Cognitive Ability	JPSP	journals	conditional	Social & Personality	JPSP	0.019	0.002	0.079	Rare	0.02, 95% CI [0.00, 0.08]
Cognitive Ability	JRP	journals	conditional	Social & Personality	JRP	0.043	0.011	0.112	Rare	0.04, 95% CI [0.01, 0.11]
Cognitive Ability	PSPB	journals	conditional	Social & Personality	PSPB	0.106	0.042	0.214	Occasional	0.11, 95% CI [0.04, 0.21]
Neuro/Bio/Psychophys	Psychology (marginal)	field	marginal	across field	across field	0.092	0.053	0.160	Uncommon	0.09, 95% CI [0.05, 0.16]
Neuro/Bio/Psychophys	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.096	0.010	0.338	Uncommon	0.10, 95% CI [0.01, 0.34]
Neuro/Bio/Psychophys	General	subfields	conditional	General	hypothetical typical journal	0.115	0.011	0.390	Occasional	0.12, 95% CI [0.01, 0.39]
Neuro/Bio/Psychophys	Collabra	journals	conditional	General	Collabra	0.278	0.177	0.404	Common	0.28, 95% CI [0.18, 0.40]
Neuro/Bio/Psychophys	JEP:G	journals	conditional	General	JEP:G	0.191	0.109	0.297	Occasional	0.19, 95% CI [0.11, 0.30]
Neuro/Bio/Psychophys	NHB	journals	conditional	General	NHB	0.129	0.071	0.208	Occasional	0.13, 95% CI [0.07, 0.21]
Neuro/Bio/Psychophys	PLOSONE	journals	conditional	General	PLOSONE	0.188	0.111	0.289	Occasional	0.19, 95% CI [0.11, 0.29]
Neuro/Bio/Psychophys	PS	journals	conditional	General	PS	0.067	0.027	0.134	Uncommon	0.07, 95% CI [0.03, 0.13]
Neuro/Bio/Psychophys	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.093	0.010	0.331	Uncommon	0.09, 95% CI [0.01, 0.33]
Neuro/Bio/Psychophys	BRAT	journals	conditional	Clinical	BRAT	0.161	0.091	0.254	Occasional	0.16, 95% CI [0.09, 0.25]
Neuro/Bio/Psychophys	CPS	journals	conditional	Clinical	CPS	0.167	0.094	0.264	Occasional	0.17, 95% CI [0.09, 0.26]
Neuro/Bio/Psychophys	JCCAP	journals	conditional	Clinical	JCCAP	0.038	0.009	0.099	Rare	0.04, 95% CI [0.01, 0.10]
Neuro/Bio/Psychophys	JCCP	journals	conditional	Clinical	JCCP	0.051	0.015	0.123	Uncommon	0.05, 95% CI [0.01, 0.12]
Neuro/Bio/Psychophys	JPCS	journals	conditional	Clinical	JPCS	0.055	0.016	0.127	Uncommon	0.06, 95% CI [0.02, 0.13]
Neuro/Bio/Psychophys	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.085	0.009	0.306	Uncommon	0.09, 95% CI [0.01, 0.31]
Neuro/Bio/Psychophys	Cognition	journals	conditional	Cognitive	Cognition	0.019	0.003	0.063	Rare	0.02, 95% CI [0.00, 0.06]
Neuro/Bio/Psychophys	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.055	0.012	0.145	Uncommon	0.05, 95% CI [0.01, 0.15]
Neuro/Bio/Psychophys	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.061	0.020	0.136	Uncommon	0.06, 95% CI [0.02, 0.14]
Neuro/Bio/Psychophys	JML	journals	conditional	Cognitive	JML	0.104	0.032	0.246	Occasional	0.10, 95% CI [0.03, 0.25]
Neuro/Bio/Psychophys	M&C	journals	conditional	Cognitive	M&C	0.234	0.133	0.364	Occasional	0.23, 95% CI [0.13, 0.36]
Neuro/Bio/Psychophys	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.082	0.008	0.304	Uncommon	0.08, 95% CI [0.01, 0.30]
Neuro/Bio/Psychophys	CD	journals	conditional	Developmental	CD	0.053	0.014	0.135	Uncommon	0.05, 95% CI [0.01, 0.13]
Neuro/Bio/Psychophys	DP	journals	conditional	Developmental	DP	0.116	0.045	0.236	Occasional	0.12, 95% CI [0.04, 0.24]
Neuro/Bio/Psychophys	JECP	journals	conditional	Developmental	JECP	0.018	0.003	0.062	Rare	0.02, 95% CI [0.00, 0.06]
Neuro/Bio/Psychophys	JRA	journals	conditional	Developmental	JRA	0.160	0.059	0.330	Occasional	0.16, 95% CI [0.06, 0.33]
Neuro/Bio/Psychophys	PA	journals	conditional	Developmental	PA	0.048	0.011	0.132	Rare	0.05, 95% CI [0.01, 0.13]
Neuro/Bio/Psychophys	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.103	0.010	0.360	Occasional	0.10, 95% CI [0.01, 0.36]
Neuro/Bio/Psychophys	JAP	journals	conditional	Industrial & Organizational	JAP	0.064	0.015	0.165	Uncommon	0.06, 95% CI [0.01, 0.17]
Neuro/Bio/Psychophys	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.147	0.059	0.294	Occasional	0.15, 95% CI [0.06, 0.29]
Neuro/Bio/Psychophys	JOB	journals	conditional	Industrial & Organizational	JOB	0.104	0.044	0.205	Occasional	0.10, 95% CI [0.04, 0.20]
Neuro/Bio/Psychophys	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.087	0.023	0.218	Uncommon	0.09, 95% CI [0.02, 0.22]
Neuro/Bio/Psychophys	PP	journals	conditional	Industrial & Organizational	PP	0.170	0.072	0.324	Occasional	0.17, 95% CI [0.07, 0.32]
Neuro/Bio/Psychophys	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.089	0.009	0.316	Uncommon	0.09, 95% CI [0.01, 0.32]
Neuro/Bio/Psychophys	Emotion	journals	conditional	Social & Personality	Emotion	0.023	0.004	0.072	Rare	0.02, 95% CI [0.00, 0.07]
Neuro/Bio/Psychophys	JESP	journals	conditional	Social & Personality	JESP	0.030	0.004	0.103	Rare	0.03, 95% CI [0.00, 0.10]
Neuro/Bio/Psychophys	JPSP	journals	conditional	Social & Personality	JPSP	0.079	0.019	0.209	Uncommon	0.08, 95% CI [0.02, 0.21]
Neuro/Bio/Psychophys	JRP	journals	conditional	Social & Personality	JRP	0.229	0.138	0.346	Occasional	0.23, 95% CI [0.14, 0.35]
Neuro/Bio/Psychophys	PSPB	journals	conditional	Social & Personality	PSPB	0.164	0.078	0.291	Occasional	0.16, 95% CI [0.08, 0.29]
Self-Report about Subjective States	Psychology (marginal)	field	marginal	across field	across field	0.868	0.676	0.944	Frequent	0.87, 95% CI [0.68, 0.94]
Self-Report about Subjective States	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.888	0.414	0.993	Frequent	0.89, 95% CI [0.41, 0.99]
Self-Report about Subjective States	General	subfields	conditional	General	hypothetical typical journal	0.973	0.827	0.997	Frequent	0.97, 95% CI [0.83, 1.00]
Self-Report about Subjective States	Collabra	journals	conditional	General	Collabra	0.991	0.964	0.999	Frequent	0.99, 95% CI [0.96, 1.00]
Self-Report about Subjective States	JEP:G	journals	conditional	General	JEP:G	0.988	0.956	0.998	Frequent	0.99, 95% CI [0.96, 1.00]
Self-Report about Subjective States	NHB	journals	conditional	General	NHB	0.883	0.805	0.937	Frequent	0.88, 95% CI [0.81, 0.94]
Self-Report about Subjective States	PLOSONE	journals	conditional	General	PLOSONE	0.961	0.904	0.989	Frequent	0.96, 95% CI [0.90, 0.99]
Self-Report about Subjective States	PS	journals	conditional	General	PS	0.988	0.958	0.998	Frequent	0.99, 95% CI [0.96, 1.00]
Self-Report about Subjective States	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.945	0.711	0.992	Frequent	0.95, 95% CI [0.71, 0.99]
Self-Report about Subjective States	BRAT	journals	conditional	Clinical	BRAT	0.947	0.889	0.980	Frequent	0.95, 95% CI [0.89, 0.98]
Self-Report about Subjective States	CPS	journals	conditional	Clinical	CPS	0.979	0.939	0.996	Frequent	0.98, 95% CI [0.94, 1.00]
Self-Report about Subjective States	JCCAP	journals	conditional	Clinical	JCCAP	0.978	0.939	0.996	Frequent	0.98, 95% CI [0.94, 1.00]
Self-Report about Subjective States	JCCP	journals	conditional	Clinical	JCCP	0.781	0.663	0.874	Frequent	0.78, 95% CI [0.66, 0.87]
Self-Report about Subjective States	JPCS	journals	conditional	Clinical	JPCS	0.940	0.874	0.977	Frequent	0.94, 95% CI [0.87, 0.98]
Self-Report about Subjective States	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.929	0.623	0.990	Frequent	0.93, 95% CI [0.62, 0.99]
Self-Report about Subjective States	Cognition	journals	conditional	Cognitive	Cognition	0.941	0.867	0.979	Frequent	0.94, 95% CI [0.87, 0.98]
Self-Report about Subjective States	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.892	0.769	0.959	Frequent	0.89, 95% CI [0.77, 0.96]
Self-Report about Subjective States	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.925	0.841	0.973	Frequent	0.93, 95% CI [0.84, 0.97]
Self-Report about Subjective States	JML	journals	conditional	Cognitive	JML	0.944	0.843	0.986	Frequent	0.94, 95% CI [0.84, 0.99]
Self-Report about Subjective States	M&C	journals	conditional	Cognitive	M&C	0.941	0.863	0.979	Frequent	0.94, 95% CI [0.86, 0.98]
Self-Report about Subjective States	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.779	0.323	0.964	Frequent	0.78, 95% CI [0.32, 0.96]
Self-Report about Subjective States	CD	journals	conditional	Developmental	CD	0.819	0.689	0.911	Frequent	0.82, 95% CI [0.69, 0.91]
Self-Report about Subjective States	DP	journals	conditional	Developmental	DP	0.811	0.670	0.913	Frequent	0.81, 95% CI [0.67, 0.91]
Self-Report about Subjective States	JECP	journals	conditional	Developmental	JECP	0.765	0.639	0.864	Frequent	0.77, 95% CI [0.64, 0.86]
Self-Report about Subjective States	JRA	journals	conditional	Developmental	JRA	0.805	0.625	0.924	Frequent	0.81, 95% CI [0.62, 0.92]
Self-Report about Subjective States	PA	journals	conditional	Developmental	PA	0.692	0.537	0.823	Frequent	0.69, 95% CI [0.54, 0.82]
Self-Report about Subjective States	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.782	0.333	0.967	Frequent	0.78, 95% CI [0.33, 0.97]
Self-Report about Subjective States	JAP	journals	conditional	Industrial & Organizational	JAP	0.802	0.656	0.908	Frequent	0.80, 95% CI [0.66, 0.91]
Self-Report about Subjective States	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.754	0.591	0.877	Frequent	0.75, 95% CI [0.59, 0.88]
Self-Report about Subjective States	JOB	journals	conditional	Industrial & Organizational	JOB	0.634	0.497	0.759	Frequent	0.63, 95% CI [0.50, 0.76]
Self-Report about Subjective States	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.854	0.702	0.948	Frequent	0.85, 95% CI [0.70, 0.95]
Self-Report about Subjective States	PP	journals	conditional	Industrial & Organizational	PP	0.563	0.393	0.727	Frequent	0.56, 95% CI [0.39, 0.73]
Self-Report about Subjective States	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.799	0.372	0.968	Frequent	0.80, 95% CI [0.37, 0.97]
Self-Report about Subjective States	Emotion	journals	conditional	Social & Personality	Emotion	0.540	0.414	0.660	Frequent	0.54, 95% CI [0.41, 0.66]
Self-Report about Subjective States	JESP	journals	conditional	Social & Personality	JESP	0.794	0.634	0.907	Frequent	0.79, 95% CI [0.63, 0.91]
Self-Report about Subjective States	JPSP	journals	conditional	Social & Personality	JPSP	0.755	0.569	0.890	Frequent	0.75, 95% CI [0.57, 0.89]
Self-Report about Subjective States	JRP	journals	conditional	Social & Personality	JRP	0.751	0.636	0.846	Frequent	0.75, 95% CI [0.64, 0.85]
Self-Report about Subjective States	PSPB	journals	conditional	Social & Personality	PSPB	0.913	0.816	0.970	Frequent	0.91, 95% CI [0.82, 0.97]
Self-Report about Behavior	Psychology (marginal)	field	marginal	across field	across field	0.073	0.040	0.152	Uncommon	0.07, 95% CI [0.04, 0.15]
Self-Report about Behavior	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.068	0.009	0.245	Uncommon	0.07, 95% CI [0.01, 0.24]
Self-Report about Behavior	General	subfields	conditional	General	hypothetical typical journal	0.043	0.006	0.161	Rare	0.04, 95% CI [0.01, 0.16]
Self-Report about Behavior	Collabra	journals	conditional	General	Collabra	0.063	0.024	0.132	Uncommon	0.06, 95% CI [0.02, 0.13]
Self-Report about Behavior	JEP:G	journals	conditional	General	JEP:G	0.080	0.030	0.177	Uncommon	0.08, 95% CI [0.03, 0.18]
Self-Report about Behavior	NHB	journals	conditional	General	NHB	0.012	0.002	0.041	Rare	0.01, 95% CI [0.00, 0.04]
Self-Report about Behavior	PLOSONE	journals	conditional	General	PLOSONE	0.023	0.004	0.068	Rare	0.02, 95% CI [0.00, 0.07]
Self-Report about Behavior	PS	journals	conditional	General	PS	0.036	0.009	0.090	Rare	0.04, 95% CI [0.01, 0.09]
Self-Report about Behavior	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.064	0.009	0.201	Uncommon	0.06, 95% CI [0.01, 0.20]
Self-Report about Behavior	BRAT	journals	conditional	Clinical	BRAT	0.113	0.061	0.191	Occasional	0.11, 95% CI [0.06, 0.19]
Self-Report about Behavior	CPS	journals	conditional	Clinical	CPS	0.048	0.014	0.107	Rare	0.05, 95% CI [0.01, 0.11]
Self-Report about Behavior	JCCAP	journals	conditional	Clinical	JCCAP	0.065	0.026	0.128	Uncommon	0.06, 95% CI [0.03, 0.13]
Self-Report about Behavior	JCCP	journals	conditional	Clinical	JCCP	0.021	0.004	0.063	Rare	0.02, 95% CI [0.00, 0.06]
Self-Report about Behavior	JPCS	journals	conditional	Clinical	JPCS	0.093	0.043	0.176	Uncommon	0.09, 95% CI [0.04, 0.18]
Self-Report about Behavior	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.087	0.012	0.270	Uncommon	0.09, 95% CI [0.01, 0.27]
Self-Report about Behavior	Cognition	journals	conditional	Cognitive	Cognition	0.110	0.052	0.203	Occasional	0.11, 95% CI [0.05, 0.20]
Self-Report about Behavior	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.102	0.039	0.212	Occasional	0.10, 95% CI [0.04, 0.21]
Self-Report about Behavior	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.055	0.015	0.126	Uncommon	0.06, 95% CI [0.02, 0.13]
Self-Report about Behavior	JML	journals	conditional	Cognitive	JML	0.095	0.034	0.212	Uncommon	0.10, 95% CI [0.03, 0.21]
Self-Report about Behavior	M&C	journals	conditional	Cognitive	M&C	0.113	0.054	0.209	Occasional	0.11, 95% CI [0.05, 0.21]
Self-Report about Behavior	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.090	0.013	0.271	Uncommon	0.09, 95% CI [0.01, 0.27]
Self-Report about Behavior	CD	journals	conditional	Developmental	CD	0.135	0.065	0.253	Occasional	0.14, 95% CI [0.07, 0.25]
Self-Report about Behavior	DP	journals	conditional	Developmental	DP	0.129	0.053	0.256	Occasional	0.13, 95% CI [0.05, 0.26]
Self-Report about Behavior	JECP	journals	conditional	Developmental	JECP	0.141	0.069	0.252	Occasional	0.14, 95% CI [0.07, 0.25]
Self-Report about Behavior	JRA	journals	conditional	Developmental	JRA	0.127	0.043	0.289	Occasional	0.13, 95% CI [0.04, 0.29]
Self-Report about Behavior	PA	journals	conditional	Developmental	PA	0.052	0.010	0.134	Uncommon	0.05, 95% CI [0.01, 0.13]
Self-Report about Behavior	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.048	0.007	0.183	Rare	0.05, 95% CI [0.01, 0.18]
Self-Report about Behavior	JAP	journals	conditional	Industrial & Organizational	JAP	0.025	0.004	0.076	Rare	0.02, 95% CI [0.00, 0.08]
Self-Report about Behavior	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.038	0.009	0.108	Rare	0.04, 95% CI [0.01, 0.11]
Self-Report about Behavior	JOB	journals	conditional	Industrial & Organizational	JOB	0.049	0.016	0.117	Rare	0.05, 95% CI [0.02, 0.12]
Self-Report about Behavior	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.056	0.015	0.153	Uncommon	0.06, 95% CI [0.02, 0.15]
Self-Report about Behavior	PP	journals	conditional	Industrial & Organizational	PP	0.023	0.004	0.074	Rare	0.02, 95% CI [0.00, 0.07]
Self-Report about Behavior	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.090	0.013	0.290	Uncommon	0.09, 95% CI [0.01, 0.29]
Self-Report about Behavior	Emotion	journals	conditional	Social & Personality	Emotion	0.059	0.021	0.125	Uncommon	0.06, 95% CI [0.02, 0.13]
Self-Report about Behavior	JESP	journals	conditional	Social & Personality	JESP	0.086	0.028	0.198	Uncommon	0.09, 95% CI [0.03, 0.20]
Self-Report about Behavior	JPSP	journals	conditional	Social & Personality	JPSP	0.160	0.067	0.336	Occasional	0.16, 95% CI [0.07, 0.34]
Self-Report about Behavior	JRP	journals	conditional	Social & Personality	JRP	0.064	0.018	0.144	Uncommon	0.06, 95% CI [0.02, 0.14]
Self-Report about Behavior	PSPB	journals	conditional	Social & Personality	PSPB	0.120	0.054	0.230	Occasional	0.12, 95% CI [0.05, 0.23]
Behavioral Proxy	Psychology (marginal)	field	marginal	across field	across field	0.053	0.027	0.104	Uncommon	0.05, 95% CI [0.03, 0.10]
Behavioral Proxy	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.049	0.004	0.304	Rare	0.05, 95% CI [0.00, 0.30]
Behavioral Proxy	General	subfields	conditional	General	hypothetical typical journal	0.047	0.004	0.301	Rare	0.05, 95% CI [0.00, 0.30]
Behavioral Proxy	Collabra	journals	conditional	General	Collabra	0.083	0.032	0.171	Uncommon	0.08, 95% CI [0.03, 0.17]
Behavioral Proxy	JEP:G	journals	conditional	General	JEP:G	0.099	0.044	0.187	Uncommon	0.10, 95% CI [0.04, 0.19]
Behavioral Proxy	NHB	journals	conditional	General	NHB	0.042	0.014	0.097	Rare	0.04, 95% CI [0.01, 0.10]
Behavioral Proxy	PLOSONE	journals	conditional	General	PLOSONE	0.012	0.001	0.045	Rare	0.01, 95% CI [0.00, 0.05]
Behavioral Proxy	PS	journals	conditional	General	PS	0.052	0.018	0.113	Uncommon	0.05, 95% CI [0.02, 0.11]
Behavioral Proxy	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.051	0.004	0.314	Uncommon	0.05, 95% CI [0.00, 0.31]
Behavioral Proxy	BRAT	journals	conditional	Clinical	BRAT	0.286	0.189	0.397	Common	0.29, 95% CI [0.19, 0.40]
Behavioral Proxy	CPS	journals	conditional	Clinical	CPS	0.047	0.016	0.106	Rare	0.05, 95% CI [0.02, 0.11]
Behavioral Proxy	JCCAP	journals	conditional	Clinical	JCCAP	0.039	0.007	0.107	Rare	0.04, 95% CI [0.01, 0.11]
Behavioral Proxy	JCCP	journals	conditional	Clinical	JCCP	0.035	0.008	0.098	Rare	0.04, 95% CI [0.01, 0.10]
Behavioral Proxy	JPCS	journals	conditional	Clinical	JPCS	0.024	0.004	0.072	Rare	0.02, 95% CI [0.00, 0.07]
Behavioral Proxy	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.052	0.004	0.318	Uncommon	0.05, 95% CI [0.00, 0.32]
Behavioral Proxy	Cognition	journals	conditional	Cognitive	Cognition	0.044	0.010	0.119	Rare	0.04, 95% CI [0.01, 0.12]
Behavioral Proxy	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.022	0.002	0.085	Rare	0.02, 95% CI [0.00, 0.09]
Behavioral Proxy	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.106	0.045	0.201	Occasional	0.11, 95% CI [0.04, 0.20]
Behavioral Proxy	JML	journals	conditional	Cognitive	JML	0.050	0.011	0.151	Uncommon	0.05, 95% CI [0.01, 0.15]
Behavioral Proxy	M&C	journals	conditional	Cognitive	M&C	0.116	0.051	0.220	Occasional	0.12, 95% CI [0.05, 0.22]
Behavioral Proxy	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.055	0.004	0.333	Uncommon	0.05, 95% CI [0.00, 0.33]
Behavioral Proxy	CD	journals	conditional	Developmental	CD	0.061	0.017	0.146	Uncommon	0.06, 95% CI [0.02, 0.15]
Behavioral Proxy	DP	journals	conditional	Developmental	DP	0.103	0.035	0.223	Occasional	0.10, 95% CI [0.04, 0.22]
Behavioral Proxy	JECP	journals	conditional	Developmental	JECP	0.019	0.002	0.068	Rare	0.02, 95% CI [0.00, 0.07]
Behavioral Proxy	JRA	journals	conditional	Developmental	JRA	0.140	0.046	0.315	Occasional	0.14, 95% CI [0.05, 0.31]
Behavioral Proxy	PA	journals	conditional	Developmental	PA	0.144	0.059	0.274	Occasional	0.14, 95% CI [0.06, 0.27]
Behavioral Proxy	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.046	0.003	0.287	Rare	0.05, 95% CI [0.00, 0.29]
Behavioral Proxy	JAP	journals	conditional	Industrial & Organizational	JAP	0.013	0.001	0.061	Rare	0.01, 95% CI [0.00, 0.06]
Behavioral Proxy	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.017	0.002	0.075	Rare	0.02, 95% CI [0.00, 0.07]
Behavioral Proxy	JOB	journals	conditional	Industrial & Organizational	JOB	0.108	0.044	0.210	Occasional	0.11, 95% CI [0.04, 0.21]
Behavioral Proxy	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.098	0.029	0.246	Uncommon	0.10, 95% CI [0.03, 0.25]
Behavioral Proxy	PP	journals	conditional	Industrial & Organizational	PP	0.019	0.002	0.079	Rare	0.02, 95% CI [0.00, 0.08]
Behavioral Proxy	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.044	0.003	0.286	Rare	0.04, 95% CI [0.00, 0.29]
Behavioral Proxy	Emotion	journals	conditional	Social & Personality	Emotion	0.022	0.003	0.075	Rare	0.02, 95% CI [0.00, 0.08]
Behavioral Proxy	JESP	journals	conditional	Social & Personality	JESP	0.018	0.002	0.080	Rare	0.02, 95% CI [0.00, 0.08]
Behavioral Proxy	JPSP	journals	conditional	Social & Personality	JPSP	0.210	0.083	0.406	Occasional	0.21, 95% CI [0.08, 0.41]
Behavioral Proxy	JRP	journals	conditional	Social & Personality	JRP	0.017	0.002	0.061	Rare	0.02, 95% CI [0.00, 0.06]
Behavioral Proxy	PSPB	journals	conditional	Social & Personality	PSPB	0.018	0.002	0.069	Rare	0.02, 95% CI [0.00, 0.07]
Direct Behavioral	Psychology (marginal)	field	marginal	across field	across field	0.014	0.005	0.069	Rare	0.01, 95% CI [0.01, 0.07]
Direct Behavioral	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.010	0.000	0.095	Rare	0.01, 95% CI [0.00, 0.09]
Direct Behavioral	General	subfields	conditional	General	hypothetical typical journal	0.008	0.000	0.076	Rare	0.01, 95% CI [0.00, 0.08]
Direct Behavioral	Collabra	journals	conditional	General	Collabra	0.044	0.012	0.113	Rare	0.04, 95% CI [0.01, 0.11]
Direct Behavioral	JEP:G	journals	conditional	General	JEP:G	0.005	0.000	0.028	Rare	0.01, 95% CI [0.00, 0.03]
Direct Behavioral	NHB	journals	conditional	General	NHB	0.003	0.000	0.021	Rare	0.00, 95% CI [0.00, 0.02]
Direct Behavioral	PLOSONE	journals	conditional	General	PLOSONE	0.002	0.000	0.016	Rare	0.00, 95% CI [0.00, 0.02]
Direct Behavioral	PS	journals	conditional	General	PS	0.008	0.000	0.041	Rare	0.01, 95% CI [0.00, 0.04]
Direct Behavioral	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.009	0.000	0.078	Rare	0.01, 95% CI [0.00, 0.08]
Direct Behavioral	BRAT	journals	conditional	Clinical	BRAT	0.041	0.013	0.099	Rare	0.04, 95% CI [0.01, 0.10]
Direct Behavioral	CPS	journals	conditional	Clinical	CPS	0.005	0.000	0.026	Rare	0.01, 95% CI [0.00, 0.03]
Direct Behavioral	JCCAP	journals	conditional	Clinical	JCCAP	0.011	0.001	0.051	Rare	0.01, 95% CI [0.00, 0.05]
Direct Behavioral	JCCP	journals	conditional	Clinical	JCCP	0.002	0.000	0.019	Rare	0.00, 95% CI [0.00, 0.02]
Direct Behavioral	JPCS	journals	conditional	Clinical	JPCS	0.006	0.000	0.031	Rare	0.01, 95% CI [0.00, 0.03]
Direct Behavioral	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.010	0.000	0.094	Rare	0.01, 95% CI [0.00, 0.09]
Direct Behavioral	Cognition	journals	conditional	Cognitive	Cognition	0.013	0.001	0.060	Rare	0.01, 95% CI [0.00, 0.06]
Direct Behavioral	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.024	0.004	0.108	Rare	0.02, 95% CI [0.00, 0.11]
Direct Behavioral	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.015	0.001	0.065	Rare	0.02, 95% CI [0.00, 0.07]
Direct Behavioral	JML	journals	conditional	Cognitive	JML	0.007	0.000	0.041	Rare	0.01, 95% CI [0.00, 0.04]
Direct Behavioral	M&C	journals	conditional	Cognitive	M&C	0.007	0.000	0.035	Rare	0.01, 95% CI [0.00, 0.03]
Direct Behavioral	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.014	0.000	0.132	Rare	0.01, 95% CI [0.00, 0.13]
Direct Behavioral	CD	journals	conditional	Developmental	CD	0.019	0.002	0.081	Rare	0.02, 95% CI [0.00, 0.08]
Direct Behavioral	DP	journals	conditional	Developmental	DP	0.022	0.003	0.090	Rare	0.02, 95% CI [0.00, 0.09]
Direct Behavioral	JECP	journals	conditional	Developmental	JECP	0.040	0.011	0.111	Rare	0.04, 95% CI [0.01, 0.11]
Direct Behavioral	JRA	journals	conditional	Developmental	JRA	0.008	0.000	0.053	Rare	0.01, 95% CI [0.00, 0.05]
Direct Behavioral	PA	journals	conditional	Developmental	PA	0.019	0.002	0.084	Rare	0.02, 95% CI [0.00, 0.08]
Direct Behavioral	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.009	0.000	0.083	Rare	0.01, 95% CI [0.00, 0.08]
Direct Behavioral	JAP	journals	conditional	Industrial & Organizational	JAP	0.004	0.000	0.025	Rare	0.00, 95% CI [0.00, 0.03]
Direct Behavioral	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.006	0.000	0.037	Rare	0.01, 95% CI [0.00, 0.04]
Direct Behavioral	JOB	journals	conditional	Industrial & Organizational	JOB	0.006	0.000	0.033	Rare	0.01, 95% CI [0.00, 0.03]
Direct Behavioral	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.022	0.003	0.108	Rare	0.02, 95% CI [0.00, 0.11]
Direct Behavioral	PP	journals	conditional	Industrial & Organizational	PP	0.007	0.000	0.043	Rare	0.01, 95% CI [0.00, 0.04]
Direct Behavioral	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.009	0.000	0.089	Rare	0.01, 95% CI [0.00, 0.09]
Direct Behavioral	Emotion	journals	conditional	Social & Personality	Emotion	0.005	0.000	0.031	Rare	0.00, 95% CI [0.00, 0.03]
Direct Behavioral	JESP	journals	conditional	Social & Personality	JESP	0.008	0.000	0.044	Rare	0.01, 95% CI [0.00, 0.04]
Direct Behavioral	JPSP	journals	conditional	Social & Personality	JPSP	0.009	0.000	0.054	Rare	0.01, 95% CI [0.00, 0.05]
Direct Behavioral	JRP	journals	conditional	Social & Personality	JRP	0.003	0.000	0.021	Rare	0.00, 95% CI [0.00, 0.02]
Direct Behavioral	PSPB	journals	conditional	Social & Personality	PSPB	0.021	0.004	0.085	Rare	0.02, 95% CI [0.00, 0.09]

Plot field + subfields

p_prevalence_subfields <- posterior_prevalence |>
  filter(re_level != "journals") |>
  ggplot(aes(prevalence_estimate, outcome)) +
  geom_rect(aes(xmin = 0.00, xmax = 0.05, ymin = -Inf, ymax = Inf, color = NULL),
            fill = "grey65") +
  geom_rect(aes(xmin = 0.05, xmax = 0.10, ymin = -Inf, ymax = Inf, color = NULL),
            fill = "grey73") +
  geom_rect(aes(xmin = 0.10, xmax = 0.25, ymin = -Inf, ymax = Inf, color = NULL),
            fill = "grey80") +
  geom_rect(aes(xmin = 0.25, xmax = 0.50, ymin = -Inf, ymax = Inf, color = NULL),
            fill = "grey90") +
  geom_rect(aes(xmin = 0.50, xmax = 1.00, ymin = -Inf, ymax = Inf, color = NULL),
            fill = "grey95") +
  annotate("text", x = 0.0250, y = "Self-Report about Behavior", label = "Rare",       size = 4, color = "grey30", angle = 90) +
  annotate("text", x = 0.0725, y = "Self-Report about Behavior", label = "Uncommon",   size = 4, color = "grey40", angle = 90) +
  annotate("text", x = 0.1750, y = "Self-Report about Behavior", label = "Occasional", size = 4, color = "grey50", angle = 90) +
  annotate("text", x = 0.3750, y = "Self-Report about Behavior", label = "Common",     size = 4, color = "grey60", angle = 90) +
  annotate("text", x = 0.7500, y = "Self-Report about Behavior", label = "Frequent",   size = 4, color = "grey65", angle = 90) +
  geom_linerangeh(aes(xmin = prevalence_ci_lower, xmax = prevalence_ci_upper), position = position_dodge(width = .7)) +
  geom_point(position = position_dodge(width = .7)) +
  coord_cartesian(xlim = c(0, 1)) +
  theme_linedraw() + 
  theme(panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        panel.spacing = unit(1, "lines")) +  # Increase spacing between facets  
  ylab("") +
  xlab("Prevalence") +
  #scale_x_continuous(labels = scales::label_percent(), breaks = c(0, 0.05, 0.1, 0.25, 0.5, 1)) +
  scale_x_continuous(labels = c("0%", "5%", "10%", "25%", "50%", "75%", "100%"),
                     breaks = c(0, 0.05, 0.10, 0.25, 0.50, 0.75, 1),
                     minor_breaks = NULL,
                     expand = c(0,0)) +
  guides(color = guide_legend(title = "Measure mode",
                              reverse = TRUE)) +
  facet_wrap(~ fct_rev(label), ncol = 2)

p_prevalence_subfields

quicksave("p_prevalence_subfields", 11, 6)

Plot journals

p_prevalence_journals <- posterior_prevalence |>
  filter(re_level == "journals") |>
  mutate(label = fct_relevel(label, subfield_and_journal_labels_without_subfields),
         label = fct_rev(label),
         outcome = fct_relevel(outcome, 
                               "Direct Behavioral",     
                               "Behavioral Proxy",
                               "Self-Report about Behavior",
                               "Self-Report about Subjective States",
                               "Neuro/Bio/Psychophys",
                               "Cognitive Ability")) |>
  ggplot(aes(prevalence_estimate, label, color = subfield)) +
  geom_linerangeh(aes(xmin = prevalence_ci_lower, xmax = prevalence_ci_upper)) +
  geom_point() +
  coord_cartesian(xlim = c(0, 1)) +
  theme_linedraw() +
  ylab("") +
  xlab("Prevalence") +
  guides(color = guide_legend(title = "Measure mode")) +
  facet_wrap(~ outcome, ncol = 3)

p_prevalence_journals

quicksave("p_prevalence_journals", 12, 10)

Trend

if(exists("models/posterior_trends.rds")){
  posterior_trends <- read_rds("models/posterior_trends.rds")
} else {
  
  est_trend <- function(subfield = "hypothetical typical subfield", journal = "hypothetical typical journal", type = c("conditional", "marginal"), fit) {
    
    # reminder of the distinction: 
    # "a typical journal..." (conditional effect, taking levels of the random effects into account, whether existing or new hypothetical ones) vs.
    # "typically, journals..." (marginal effect, ignoring the random effects, therefore likely producing wider estimates given any heterogeneity among the random effects)
    
    if(type == "conditional") {
      # conditional effect for either existing levels of the random effects if specified,
      # or if not specified then for hypothetical new representative levels 
      if(subfield == "hypothetical typical subfield"){
        subfield = factor("hypothetical typical subfield", levels = unique(fit$data$subfield))
      }
      
      if(journal == "hypothetical typical journal"){
        journal = factor("hypothetical typical journal", levels = unique(fit$data$journal))
      }
      
      res <- 
        slopes(model = fit,
               variables = "year_centered", 
               newdata = datagrid(subfield = subfield,
                                  journal = journal),
               re_formula = NULL, 
               allow_new_levels = TRUE) |>
        select(outcome = group,
               trend_estimate = estimate,
               trend_ci_lower = conf.low,
               trend_ci_upper = conf.high)
      
    } else if (type == "marginal") {
      res <- 
        avg_slopes(model = fit,
                   variables = "year_centered",
                   re_formula = NA, 
                   allow_new_levels = FALSE) |>
        select(outcome = group,
               trend_estimate = estimate,
               trend_ci_lower = conf.low,
               trend_ci_upper = conf.high)
    }
    
    return(res)
  }
  
  # field - marginal effect
  posterior_trend_field_marginal <- 
    expand_grid(subfield = "across field", 
                journal = "across field",
                type = "marginal",
                re_level = "field") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_trend, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # field - conditional effect
  posterior_trend_field_conditional <- 
    expand_grid(subfield = "hypothetical typical subfield", 
                journal = "hypothetical typical journal",
                type = "conditional",
                re_level = "field") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_trend, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # subfields
  posterior_trend_subfields <- 
    expand_grid(data_centered |>
                  distinct(subfield),
                journal = "hypothetical typical journal",
                type = "conditional",
                re_level = "subfields") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_trend, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # journals
  posterior_trend_journals <- 
    expand_grid(data_centered |>
                  distinct(subfield, journal),
                type = "conditional",
                re_level = "journals") |>
    tibble(res = pmap(list(subfield, journal, type), 
                      est_trend, 
                      fit = fitted_model)) |>
    unnest(res) 
  
  # combine
  posterior_trends <- 
    bind_rows(posterior_trend_field_marginal,
              posterior_trend_field_conditional,
              posterior_trend_subfields,
              posterior_trend_journals) |>
    mutate(subfield = ifelse(is.na(subfield), "", subfield),
           journal = ifelse(is.na(journal), "", journal)) |>
    mutate(label = case_when(re_level == "field" & type == "marginal" ~ "Psychology (marginal)",
                             re_level == "field" & type == "conditional" ~ "Psychology (conditional)",
                             re_level == "subfields" & journal == "hypothetical typical journal" ~ subfield,
                             journal != "hypothetical typical journal" & !is.na(journal) ~ journal)) |>
    mutate(label = fct_relevel(label,
                               "Psychology (marginal)",
                               "Psychology (conditional)",
                               subfield_and_journal_labels),
    label = fct_rev(label)) |>
    mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",     
                               outcome == "behavioralproxy" ~ "Behavioral Proxy",
                               outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                               outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                               outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                               outcome == "cognitiveability" ~ "Cognitive Ability",
                               TRUE ~ outcome),
           outcome = fct_relevel(outcome, 
                                 "Direct Behavioral",       
                                 "Behavioral Proxy",
                                 "Self-Report about Behavior",
                                 "Self-Report about Subjective States",
                                 "Neuro/Bio/Psychophys",
                                 "Cognitive Ability"),
           outcome = fct_rev(outcome),
           subfield = fct_relevel(subfield,
                                  "hypothetical typical subfield",
                                  "General",
                                  "Clinical",
                                  "Cognitive",
                                  "Developmental",
                                  "Industrial & Organizational",
                                  "Social & Personality")) |>
    mutate(trend_category = case_when(trend_ci_lower > 0 ~ "Increasing",
                                      trend_ci_upper < 0 ~ "Decreasing",
                                      TRUE ~ "No trend detected"),
           trend_results = paste0(rnd(trend_estimate), ", 95% CI [", rnd(trend_ci_lower), ", ", rnd(trend_ci_upper), "]")) |>
    select(outcome, label, re_level, type, subfield, journal, 
           trend_estimate, trend_ci_lower, trend_ci_upper, trend_category, trend_results) |>
    arrange(outcome, desc(label), subfield)
  
  write_rds(posterior_trends, "models/posterior_trends.rds")
}

# print table
posterior_trends |>
  mutate_if(is.numeric, round_half_up, digits = 3) |>
  kable() |>
  kable_classic(full_width = FALSE)

outcome	label	re_level	type	subfield	journal	trend_estimate	trend_ci_lower	trend_ci_upper	trend_category	trend_results
Cognitive Ability	Psychology (marginal)	field	marginal	across field	across field	0.002	-0.004	0.009	No trend detected	0.00, 95% CI [-0.00, 0.01]
Cognitive Ability	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.001	-0.029	0.028	No trend detected	0.00, 95% CI [-0.03, 0.03]
Cognitive Ability	General	subfields	conditional	General	hypothetical typical journal	0.000	-0.034	0.024	No trend detected	0.00, 95% CI [-0.03, 0.02]
Cognitive Ability	Collabra	journals	conditional	General	Collabra	0.004	-0.006	0.018	No trend detected	0.00, 95% CI [-0.01, 0.02]
Cognitive Ability	JEP:G	journals	conditional	General	JEP:G	-0.002	-0.012	0.011	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	NHB	journals	conditional	General	NHB	-0.009	-0.020	-0.003	Decreasing	-0.01, 95% CI [-0.02, -0.00]
Cognitive Ability	PLOSONE	journals	conditional	General	PLOSONE	0.025	0.010	0.047	Increasing	0.03, 95% CI [0.01, 0.05]
Cognitive Ability	PS	journals	conditional	General	PS	0.001	-0.003	0.007	No trend detected	0.00, 95% CI [-0.00, 0.01]
Cognitive Ability	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.003	-0.025	0.031	No trend detected	0.00, 95% CI [-0.03, 0.03]
Cognitive Ability	BRAT	journals	conditional	Clinical	BRAT	0.001	-0.004	0.005	No trend detected	0.00, 95% CI [-0.00, 0.01]
Cognitive Ability	CPS	journals	conditional	Clinical	CPS	0.003	-0.011	0.015	No trend detected	0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	JCCAP	journals	conditional	Clinical	JCCAP	0.006	0.002	0.013	Increasing	0.01, 95% CI [0.00, 0.01]
Cognitive Ability	JCCP	journals	conditional	Clinical	JCCP	0.007	-0.001	0.018	No trend detected	0.01, 95% CI [-0.00, 0.02]
Cognitive Ability	JPCS	journals	conditional	Clinical	JPCS	0.011	-0.002	0.023	No trend detected	0.01, 95% CI [-0.00, 0.02]
Cognitive Ability	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.002	-0.026	0.031	No trend detected	0.00, 95% CI [-0.03, 0.03]
Cognitive Ability	Cognition	journals	conditional	Cognitive	Cognition	0.011	0.002	0.026	Increasing	0.01, 95% CI [0.00, 0.03]
Cognitive Ability	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	-0.002	-0.018	0.014	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Cognitive Ability	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	-0.001	-0.008	0.005	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	JML	journals	conditional	Cognitive	JML	0.008	-0.010	0.032	No trend detected	0.01, 95% CI [-0.01, 0.03]
Cognitive Ability	M&C	journals	conditional	Cognitive	M&C	0.003	-0.014	0.017	No trend detected	0.00, 95% CI [-0.01, 0.02]
Cognitive Ability	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.001	-0.032	0.026	No trend detected	0.00, 95% CI [-0.03, 0.03]
Cognitive Ability	CD	journals	conditional	Developmental	CD	0.004	-0.005	0.015	No trend detected	0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	DP	journals	conditional	Developmental	DP	0.000	-0.009	0.007	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	JECP	journals	conditional	Developmental	JECP	-0.008	-0.020	0.000	Decreasing	-0.01, 95% CI [-0.02, -0.00]
Cognitive Ability	JRA	journals	conditional	Developmental	JRA	0.005	-0.015	0.023	No trend detected	0.00, 95% CI [-0.01, 0.02]
Cognitive Ability	PA	journals	conditional	Developmental	PA	0.005	-0.004	0.017	No trend detected	0.00, 95% CI [-0.00, 0.02]
Cognitive Ability	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.001	-0.031	0.029	No trend detected	0.00, 95% CI [-0.03, 0.03]
Cognitive Ability	JAP	journals	conditional	Industrial & Organizational	JAP	-0.009	-0.029	0.013	No trend detected	-0.01, 95% CI [-0.03, 0.01]
Cognitive Ability	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	-0.006	-0.036	0.016	No trend detected	-0.01, 95% CI [-0.04, 0.02]
Cognitive Ability	JOB	journals	conditional	Industrial & Organizational	JOB	0.001	-0.005	0.008	No trend detected	0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.005	-0.007	0.021	No trend detected	0.01, 95% CI [-0.01, 0.02]
Cognitive Ability	PP	journals	conditional	Industrial & Organizational	PP	-0.008	-0.034	0.008	No trend detected	-0.01, 95% CI [-0.03, 0.01]
Cognitive Ability	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.002	-0.030	0.030	No trend detected	0.00, 95% CI [-0.03, 0.03]
Cognitive Ability	Emotion	journals	conditional	Social & Personality	Emotion	-0.031	-0.059	-0.008	Decreasing	-0.03, 95% CI [-0.06, -0.01]
Cognitive Ability	JESP	journals	conditional	Social & Personality	JESP	0.000	-0.013	0.009	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Cognitive Ability	JPSP	journals	conditional	Social & Personality	JPSP	0.001	-0.006	0.015	No trend detected	0.00, 95% CI [-0.01, 0.02]
Cognitive Ability	JRP	journals	conditional	Social & Personality	JRP	0.009	0.002	0.020	Increasing	0.01, 95% CI [0.00, 0.02]
Cognitive Ability	PSPB	journals	conditional	Social & Personality	PSPB	-0.001	-0.018	0.021	No trend detected	-0.00, 95% CI [-0.02, 0.02]
Neuro/Bio/Psychophys	Psychology (marginal)	field	marginal	across field	across field	0.002	-0.008	0.012	No trend detected	0.00, 95% CI [-0.01, 0.01]
Neuro/Bio/Psychophys	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.002	-0.029	0.037	No trend detected	0.00, 95% CI [-0.03, 0.04]
Neuro/Bio/Psychophys	General	subfields	conditional	General	hypothetical typical journal	0.004	-0.030	0.044	No trend detected	0.00, 95% CI [-0.03, 0.04]
Neuro/Bio/Psychophys	Collabra	journals	conditional	General	Collabra	0.022	-0.001	0.051	No trend detected	0.02, 95% CI [-0.00, 0.05]
Neuro/Bio/Psychophys	JEP:G	journals	conditional	General	JEP:G	0.000	-0.022	0.026	No trend detected	0.00, 95% CI [-0.02, 0.03]
Neuro/Bio/Psychophys	NHB	journals	conditional	General	NHB	0.020	0.001	0.044	Increasing	0.02, 95% CI [0.00, 0.04]
Neuro/Bio/Psychophys	PLOSONE	journals	conditional	General	PLOSONE	0.026	0.004	0.051	Increasing	0.03, 95% CI [0.00, 0.05]
Neuro/Bio/Psychophys	PS	journals	conditional	General	PS	-0.003	-0.014	0.009	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Neuro/Bio/Psychophys	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.004	-0.026	0.038	No trend detected	0.00, 95% CI [-0.03, 0.04]
Neuro/Bio/Psychophys	BRAT	journals	conditional	Clinical	BRAT	0.019	0.003	0.037	Increasing	0.02, 95% CI [0.00, 0.04]
Neuro/Bio/Psychophys	CPS	journals	conditional	Clinical	CPS	0.015	0.000	0.030	Increasing	0.01, 95% CI [0.00, 0.03]
Neuro/Bio/Psychophys	JCCAP	journals	conditional	Clinical	JCCAP	0.006	0.001	0.013	Increasing	0.01, 95% CI [0.00, 0.01]
Neuro/Bio/Psychophys	JCCP	journals	conditional	Clinical	JCCP	0.008	0.000	0.021	Increasing	0.01, 95% CI [0.00, 0.02]
Neuro/Bio/Psychophys	JPCS	journals	conditional	Clinical	JPCS	-0.013	-0.030	-0.002	Decreasing	-0.01, 95% CI [-0.03, -0.00]
Neuro/Bio/Psychophys	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.002	-0.027	0.035	No trend detected	0.00, 95% CI [-0.03, 0.03]
Neuro/Bio/Psychophys	Cognition	journals	conditional	Cognitive	Cognition	0.000	-0.006	0.008	No trend detected	0.00, 95% CI [-0.01, 0.01]
Neuro/Bio/Psychophys	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	-0.008	-0.025	0.004	No trend detected	-0.01, 95% CI [-0.03, 0.00]
Neuro/Bio/Psychophys	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.006	-0.006	0.020	No trend detected	0.01, 95% CI [-0.01, 0.02]
Neuro/Bio/Psychophys	JML	journals	conditional	Cognitive	JML	0.014	-0.005	0.042	No trend detected	0.01, 95% CI [-0.01, 0.04]
Neuro/Bio/Psychophys	M&C	journals	conditional	Cognitive	M&C	0.003	-0.036	0.036	No trend detected	0.00, 95% CI [-0.04, 0.04]
Neuro/Bio/Psychophys	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.000	-0.032	0.030	No trend detected	0.00, 95% CI [-0.03, 0.03]
Neuro/Bio/Psychophys	CD	journals	conditional	Developmental	CD	0.005	-0.006	0.017	No trend detected	0.01, 95% CI [-0.01, 0.02]
Neuro/Bio/Psychophys	DP	journals	conditional	Developmental	DP	-0.011	-0.033	0.010	No trend detected	-0.01, 95% CI [-0.03, 0.01]
Neuro/Bio/Psychophys	JECP	journals	conditional	Developmental	JECP	0.000	-0.006	0.006	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Neuro/Bio/Psychophys	JRA	journals	conditional	Developmental	JRA	-0.011	-0.045	0.015	No trend detected	-0.01, 95% CI [-0.05, 0.01]
Neuro/Bio/Psychophys	PA	journals	conditional	Developmental	PA	-0.007	-0.022	0.003	No trend detected	-0.01, 95% CI [-0.02, 0.00]
Neuro/Bio/Psychophys	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.001	-0.032	0.037	No trend detected	0.00, 95% CI [-0.03, 0.04]
Neuro/Bio/Psychophys	JAP	journals	conditional	Industrial & Organizational	JAP	0.010	-0.001	0.030	No trend detected	0.01, 95% CI [-0.00, 0.03]
Neuro/Bio/Psychophys	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	-0.009	-0.042	0.017	No trend detected	-0.01, 95% CI [-0.04, 0.02]
Neuro/Bio/Psychophys	JOB	journals	conditional	Industrial & Organizational	JOB	-0.005	-0.023	0.012	No trend detected	-0.01, 95% CI [-0.02, 0.01]
Neuro/Bio/Psychophys	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.000	-0.031	0.022	No trend detected	0.00, 95% CI [-0.03, 0.02]
Neuro/Bio/Psychophys	PP	journals	conditional	Industrial & Organizational	PP	-0.007	-0.045	0.022	No trend detected	-0.01, 95% CI [-0.05, 0.02]
Neuro/Bio/Psychophys	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.002	-0.026	0.036	No trend detected	0.00, 95% CI [-0.03, 0.04]
Neuro/Bio/Psychophys	Emotion	journals	conditional	Social & Personality	Emotion	0.000	-0.009	0.006	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Neuro/Bio/Psychophys	JESP	journals	conditional	Social & Personality	JESP	0.001	-0.010	0.011	No trend detected	0.00, 95% CI [-0.01, 0.01]
Neuro/Bio/Psychophys	JPSP	journals	conditional	Social & Personality	JPSP	0.010	-0.008	0.047	No trend detected	0.01, 95% CI [-0.01, 0.05]
Neuro/Bio/Psychophys	JRP	journals	conditional	Social & Personality	JRP	0.003	-0.017	0.022	No trend detected	0.00, 95% CI [-0.02, 0.02]
Neuro/Bio/Psychophys	PSPB	journals	conditional	Social & Personality	PSPB	-0.001	-0.023	0.024	No trend detected	-0.00, 95% CI [-0.02, 0.02]
Self-Report about Subjective States	Psychology (marginal)	field	marginal	across field	across field	0.005	-0.007	0.019	No trend detected	0.00, 95% CI [-0.01, 0.02]
Self-Report about Subjective States	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.002	-0.022	0.038	No trend detected	0.00, 95% CI [-0.02, 0.04]
Self-Report about Subjective States	General	subfields	conditional	General	hypothetical typical journal	0.001	-0.010	0.011	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Subjective States	Collabra	journals	conditional	General	Collabra	0.000	-0.003	0.003	No trend detected	0.00, 95% CI [-0.00, 0.00]
Self-Report about Subjective States	JEP:G	journals	conditional	General	JEP:G	0.001	-0.002	0.005	No trend detected	0.00, 95% CI [-0.00, 0.00]
Self-Report about Subjective States	NHB	journals	conditional	General	NHB	-0.005	-0.026	0.009	No trend detected	-0.00, 95% CI [-0.03, 0.01]
Self-Report about Subjective States	PLOSONE	journals	conditional	General	PLOSONE	0.005	-0.002	0.014	No trend detected	0.00, 95% CI [-0.00, 0.01]
Self-Report about Subjective States	PS	journals	conditional	General	PS	0.000	-0.003	0.003	No trend detected	0.00, 95% CI [-0.00, 0.00]
Self-Report about Subjective States	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.001	-0.015	0.018	No trend detected	0.00, 95% CI [-0.01, 0.02]
Self-Report about Subjective States	BRAT	journals	conditional	Clinical	BRAT	0.003	-0.004	0.013	No trend detected	0.00, 95% CI [-0.00, 0.01]
Self-Report about Subjective States	CPS	journals	conditional	Clinical	CPS	0.000	-0.003	0.005	No trend detected	0.00, 95% CI [-0.00, 0.00]
Self-Report about Subjective States	JCCAP	journals	conditional	Clinical	JCCAP	0.001	-0.002	0.007	No trend detected	0.00, 95% CI [-0.00, 0.01]
Self-Report about Subjective States	JCCP	journals	conditional	Clinical	JCCP	0.002	-0.016	0.019	No trend detected	0.00, 95% CI [-0.02, 0.02]
Self-Report about Subjective States	JPCS	journals	conditional	Clinical	JPCS	0.001	-0.008	0.010	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Subjective States	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.000	-0.027	0.018	No trend detected	-0.00, 95% CI [-0.03, 0.02]
Self-Report about Subjective States	Cognition	journals	conditional	Cognitive	Cognition	-0.001	-0.013	0.009	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Self-Report about Subjective States	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	-0.003	-0.025	0.012	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Self-Report about Subjective States	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	-0.006	-0.021	0.004	No trend detected	-0.01, 95% CI [-0.02, 0.00]
Self-Report about Subjective States	JML	journals	conditional	Cognitive	JML	0.002	-0.010	0.017	No trend detected	0.00, 95% CI [-0.01, 0.02]
Self-Report about Subjective States	M&C	journals	conditional	Cognitive	M&C	0.000	-0.011	0.014	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Subjective States	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.013	-0.019	0.052	No trend detected	0.01, 95% CI [-0.02, 0.05]
Self-Report about Subjective States	CD	journals	conditional	Developmental	CD	0.018	0.001	0.043	Increasing	0.02, 95% CI [0.00, 0.04]
Self-Report about Subjective States	DP	journals	conditional	Developmental	DP	0.019	0.001	0.045	Increasing	0.02, 95% CI [0.00, 0.05]
Self-Report about Subjective States	JECP	journals	conditional	Developmental	JECP	0.018	-0.002	0.039	No trend detected	0.02, 95% CI [-0.00, 0.04]
Self-Report about Subjective States	JRA	journals	conditional	Developmental	JRA	0.026	0.004	0.067	Increasing	0.03, 95% CI [0.00, 0.07]
Self-Report about Subjective States	PA	journals	conditional	Developmental	PA	0.006	-0.025	0.033	No trend detected	0.01, 95% CI [-0.03, 0.03]
Self-Report about Subjective States	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.005	-0.030	0.041	No trend detected	0.00, 95% CI [-0.03, 0.04]
Self-Report about Subjective States	JAP	journals	conditional	Industrial & Organizational	JAP	0.007	-0.012	0.028	No trend detected	0.01, 95% CI [-0.01, 0.03]
Self-Report about Subjective States	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.002	-0.026	0.028	No trend detected	0.00, 95% CI [-0.03, 0.03]
Self-Report about Subjective States	JOB	journals	conditional	Industrial & Organizational	JOB	0.009	-0.015	0.034	No trend detected	0.01, 95% CI [-0.01, 0.03]
Self-Report about Subjective States	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.002	-0.022	0.025	No trend detected	0.00, 95% CI [-0.02, 0.02]
Self-Report about Subjective States	PP	journals	conditional	Industrial & Organizational	PP	-0.003	-0.043	0.028	No trend detected	-0.00, 95% CI [-0.04, 0.03]
Self-Report about Subjective States	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.006	-0.026	0.041	No trend detected	0.01, 95% CI [-0.03, 0.04]
Self-Report about Subjective States	Emotion	journals	conditional	Social & Personality	Emotion	0.009	-0.014	0.033	No trend detected	0.01, 95% CI [-0.01, 0.03]
Self-Report about Subjective States	JESP	journals	conditional	Social & Personality	JESP	0.006	-0.015	0.032	No trend detected	0.01, 95% CI [-0.01, 0.03]
Self-Report about Subjective States	JPSP	journals	conditional	Social & Personality	JPSP	0.020	-0.005	0.057	No trend detected	0.02, 95% CI [-0.00, 0.06]
Self-Report about Subjective States	JRP	journals	conditional	Social & Personality	JRP	0.003	-0.016	0.020	No trend detected	0.00, 95% CI [-0.02, 0.02]
Self-Report about Subjective States	PSPB	journals	conditional	Social & Personality	PSPB	0.003	-0.010	0.015	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	Psychology (marginal)	field	marginal	across field	across field	-0.001	-0.013	0.010	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.000	-0.049	0.032	No trend detected	-0.00, 95% CI [-0.05, 0.03]
Self-Report about Behavior	General	subfields	conditional	General	hypothetical typical journal	-0.003	-0.044	0.016	No trend detected	-0.00, 95% CI [-0.04, 0.02]
Self-Report about Behavior	Collabra	journals	conditional	General	Collabra	0.008	-0.003	0.024	No trend detected	0.01, 95% CI [-0.00, 0.02]
Self-Report about Behavior	JEP:G	journals	conditional	General	JEP:G	-0.036	-0.058	-0.020	Decreasing	-0.04, 95% CI [-0.06, -0.02]
Self-Report about Behavior	NHB	journals	conditional	General	NHB	0.000	-0.004	0.007	No trend detected	0.00, 95% CI [-0.00, 0.01]
Self-Report about Behavior	PLOSONE	journals	conditional	General	PLOSONE	-0.008	-0.018	-0.003	Decreasing	-0.01, 95% CI [-0.02, -0.00]
Self-Report about Behavior	PS	journals	conditional	General	PS	-0.010	-0.020	-0.003	Decreasing	-0.01, 95% CI [-0.02, -0.00]
Self-Report about Behavior	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.002	-0.042	0.033	No trend detected	0.00, 95% CI [-0.04, 0.03]
Self-Report about Behavior	BRAT	journals	conditional	Clinical	BRAT	0.002	-0.013	0.018	No trend detected	0.00, 95% CI [-0.01, 0.02]
Self-Report about Behavior	CPS	journals	conditional	Clinical	CPS	0.008	0.002	0.017	Increasing	0.01, 95% CI [0.00, 0.02]
Self-Report about Behavior	JCCAP	journals	conditional	Clinical	JCCAP	0.001	-0.012	0.010	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	JCCP	journals	conditional	Clinical	JCCP	0.001	-0.005	0.008	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	JPCS	journals	conditional	Clinical	JPCS	0.004	-0.012	0.017	No trend detected	0.00, 95% CI [-0.01, 0.02]
Self-Report about Behavior	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.000	-0.056	0.038	No trend detected	0.00, 95% CI [-0.06, 0.04]
Self-Report about Behavior	Cognition	journals	conditional	Cognitive	Cognition	-0.003	-0.021	0.018	No trend detected	-0.00, 95% CI [-0.02, 0.02]
Self-Report about Behavior	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	-0.004	-0.027	0.019	No trend detected	-0.00, 95% CI [-0.03, 0.02]
Self-Report about Behavior	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.010	-0.001	0.025	No trend detected	0.01, 95% CI [-0.00, 0.02]
Self-Report about Behavior	JML	journals	conditional	Cognitive	JML	0.006	-0.017	0.030	No trend detected	0.01, 95% CI [-0.02, 0.03]
Self-Report about Behavior	M&C	journals	conditional	Cognitive	M&C	-0.011	-0.045	0.014	No trend detected	-0.01, 95% CI [-0.05, 0.01]
Self-Report about Behavior	Developmental	subfields	conditional	Developmental	hypothetical typical journal	-0.002	-0.059	0.031	No trend detected	-0.00, 95% CI [-0.06, 0.03]
Self-Report about Behavior	CD	journals	conditional	Developmental	CD	0.009	-0.012	0.027	No trend detected	0.01, 95% CI [-0.01, 0.03]
Self-Report about Behavior	DP	journals	conditional	Developmental	DP	-0.026	-0.054	-0.004	Decreasing	-0.03, 95% CI [-0.05, -0.00]
Self-Report about Behavior	JECP	journals	conditional	Developmental	JECP	0.022	0.003	0.044	Increasing	0.02, 95% CI [0.00, 0.04]
Self-Report about Behavior	JRA	journals	conditional	Developmental	JRA	-0.041	-0.093	-0.013	Decreasing	-0.04, 95% CI [-0.09, -0.01]
Self-Report about Behavior	PA	journals	conditional	Developmental	PA	-0.001	-0.016	0.012	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Self-Report about Behavior	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.001	-0.037	0.027	No trend detected	0.00, 95% CI [-0.04, 0.03]
Self-Report about Behavior	JAP	journals	conditional	Industrial & Organizational	JAP	0.002	-0.005	0.012	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.007	-0.003	0.023	No trend detected	0.01, 95% CI [-0.00, 0.02]
Self-Report about Behavior	JOB	journals	conditional	Industrial & Organizational	JOB	-0.002	-0.015	0.010	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.001	-0.027	0.019	No trend detected	0.00, 95% CI [-0.03, 0.02]
Self-Report about Behavior	PP	journals	conditional	Industrial & Organizational	PP	0.000	-0.012	0.010	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	-0.001	-0.058	0.036	No trend detected	-0.00, 95% CI [-0.06, 0.04]
Self-Report about Behavior	Emotion	journals	conditional	Social & Personality	Emotion	0.004	-0.009	0.014	No trend detected	0.00, 95% CI [-0.01, 0.01]
Self-Report about Behavior	JESP	journals	conditional	Social & Personality	JESP	0.011	-0.009	0.031	No trend detected	0.01, 95% CI [-0.01, 0.03]
Self-Report about Behavior	JPSP	journals	conditional	Social & Personality	JPSP	-0.014	-0.046	0.025	No trend detected	-0.01, 95% CI [-0.05, 0.02]
Self-Report about Behavior	JRP	journals	conditional	Social & Personality	JRP	-0.016	-0.029	-0.006	Decreasing	-0.02, 95% CI [-0.03, -0.01]
Self-Report about Behavior	PSPB	journals	conditional	Social & Personality	PSPB	0.005	-0.015	0.029	No trend detected	0.00, 95% CI [-0.01, 0.03]
Behavioral Proxy	Psychology (marginal)	field	marginal	across field	across field	0.002	-0.008	0.014	No trend detected	0.00, 95% CI [-0.01, 0.01]
Behavioral Proxy	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.001	-0.026	0.036	No trend detected	0.00, 95% CI [-0.03, 0.04]
Behavioral Proxy	General	subfields	conditional	General	hypothetical typical journal	-0.003	-0.039	0.016	No trend detected	-0.00, 95% CI [-0.04, 0.02]
Behavioral Proxy	Collabra	journals	conditional	General	Collabra	-0.010	-0.021	0.003	No trend detected	-0.01, 95% CI [-0.02, 0.00]
Behavioral Proxy	JEP:G	journals	conditional	General	JEP:G	-0.009	-0.023	0.008	No trend detected	-0.01, 95% CI [-0.02, 0.01]
Behavioral Proxy	NHB	journals	conditional	General	NHB	-0.008	-0.017	0.000	Decreasing	-0.01, 95% CI [-0.02, -0.00]
Behavioral Proxy	PLOSONE	journals	conditional	General	PLOSONE	0.000	-0.005	0.004	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Behavioral Proxy	PS	journals	conditional	General	PS	-0.006	-0.016	0.003	No trend detected	-0.01, 95% CI [-0.02, 0.00]
Behavioral Proxy	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.008	-0.007	0.052	No trend detected	0.01, 95% CI [-0.01, 0.05]
Behavioral Proxy	BRAT	journals	conditional	Clinical	BRAT	0.037	0.015	0.062	Increasing	0.04, 95% CI [0.01, 0.06]
Behavioral Proxy	CPS	journals	conditional	Clinical	CPS	0.004	-0.004	0.013	No trend detected	0.00, 95% CI [-0.00, 0.01]
Behavioral Proxy	JCCAP	journals	conditional	Clinical	JCCAP	0.012	0.004	0.022	Increasing	0.01, 95% CI [0.00, 0.02]
Behavioral Proxy	JCCP	journals	conditional	Clinical	JCCP	0.008	0.002	0.020	Increasing	0.01, 95% CI [0.00, 0.02]
Behavioral Proxy	JPCS	journals	conditional	Clinical	JPCS	0.007	0.002	0.016	Increasing	0.01, 95% CI [0.00, 0.02]
Behavioral Proxy	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.002	-0.020	0.032	No trend detected	0.00, 95% CI [-0.02, 0.03]
Behavioral Proxy	Cognition	journals	conditional	Cognitive	Cognition	0.010	0.001	0.026	Increasing	0.01, 95% CI [0.00, 0.03]
Behavioral Proxy	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	0.002	-0.004	0.013	No trend detected	0.00, 95% CI [-0.00, 0.01]
Behavioral Proxy	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	0.006	-0.010	0.023	No trend detected	0.01, 95% CI [-0.01, 0.02]
Behavioral Proxy	JML	journals	conditional	Cognitive	JML	0.000	-0.016	0.014	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Behavioral Proxy	M&C	journals	conditional	Cognitive	M&C	-0.005	-0.039	0.015	No trend detected	-0.00, 95% CI [-0.04, 0.02]
Behavioral Proxy	Developmental	subfields	conditional	Developmental	hypothetical typical journal	-0.001	-0.036	0.020	No trend detected	-0.00, 95% CI [-0.04, 0.02]
Behavioral Proxy	CD	journals	conditional	Developmental	CD	-0.003	-0.018	0.008	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Behavioral Proxy	DP	journals	conditional	Developmental	DP	-0.011	-0.032	0.007	No trend detected	-0.01, 95% CI [-0.03, 0.01]
Behavioral Proxy	JECP	journals	conditional	Developmental	JECP	0.000	-0.005	0.007	No trend detected	0.00, 95% CI [-0.01, 0.01]
Behavioral Proxy	JRA	journals	conditional	Developmental	JRA	-0.001	-0.028	0.021	No trend detected	-0.00, 95% CI [-0.03, 0.02]
Behavioral Proxy	PA	journals	conditional	Developmental	PA	0.007	-0.014	0.029	No trend detected	0.01, 95% CI [-0.01, 0.03]
Behavioral Proxy	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.001	-0.024	0.026	No trend detected	0.00, 95% CI [-0.02, 0.03]
Behavioral Proxy	JAP	journals	conditional	Industrial & Organizational	JAP	0.000	-0.004	0.007	No trend detected	0.00, 95% CI [-0.00, 0.01]
Behavioral Proxy	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	0.000	-0.008	0.007	No trend detected	0.00, 95% CI [-0.01, 0.01]
Behavioral Proxy	JOB	journals	conditional	Industrial & Organizational	JOB	0.009	-0.007	0.028	No trend detected	0.01, 95% CI [-0.01, 0.03]
Behavioral Proxy	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	-0.001	-0.040	0.020	No trend detected	-0.00, 95% CI [-0.04, 0.02]
Behavioral Proxy	PP	journals	conditional	Industrial & Organizational	PP	0.000	-0.010	0.007	No trend detected	0.00, 95% CI [-0.01, 0.01]
Behavioral Proxy	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.003	-0.015	0.034	No trend detected	0.00, 95% CI [-0.02, 0.03]
Behavioral Proxy	Emotion	journals	conditional	Social & Personality	Emotion	0.005	0.001	0.013	Increasing	0.01, 95% CI [0.00, 0.01]
Behavioral Proxy	JESP	journals	conditional	Social & Personality	JESP	0.002	-0.005	0.010	No trend detected	0.00, 95% CI [-0.00, 0.01]
Behavioral Proxy	JPSP	journals	conditional	Social & Personality	JPSP	0.010	-0.025	0.053	No trend detected	0.01, 95% CI [-0.03, 0.05]
Behavioral Proxy	JRP	journals	conditional	Social & Personality	JRP	0.001	-0.004	0.007	No trend detected	0.00, 95% CI [-0.00, 0.01]
Behavioral Proxy	PSPB	journals	conditional	Social & Personality	PSPB	0.001	-0.004	0.009	No trend detected	0.00, 95% CI [-0.00, 0.01]
Direct Behavioral	Psychology (marginal)	field	marginal	across field	across field	-0.001	-0.009	0.003	No trend detected	-0.00, 95% CI [-0.01, 0.00]
Direct Behavioral	Psychology (conditional)	field	conditional	hypothetical typical subfield	hypothetical typical journal	0.000	-0.018	0.015	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Direct Behavioral	General	subfields	conditional	General	hypothetical typical journal	0.000	-0.016	0.014	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Direct Behavioral	Collabra	journals	conditional	General	Collabra	0.002	-0.007	0.016	No trend detected	0.00, 95% CI [-0.01, 0.02]
Direct Behavioral	JEP:G	journals	conditional	General	JEP:G	0.000	-0.003	0.005	No trend detected	0.00, 95% CI [-0.00, 0.01]
Direct Behavioral	NHB	journals	conditional	General	NHB	-0.002	-0.011	0.000	Decreasing	-0.00, 95% CI [-0.01, -0.00]
Direct Behavioral	PLOSONE	journals	conditional	General	PLOSONE	0.000	-0.002	0.003	No trend detected	0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	PS	journals	conditional	General	PS	0.005	0.000	0.015	Increasing	0.00, 95% CI [0.00, 0.01]
Direct Behavioral	Clinical	subfields	conditional	Clinical	hypothetical typical journal	0.000	-0.015	0.013	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	BRAT	journals	conditional	Clinical	BRAT	-0.003	-0.014	0.006	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	CPS	journals	conditional	Clinical	CPS	0.000	-0.004	0.003	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	JCCAP	journals	conditional	Clinical	JCCAP	0.005	0.001	0.015	Increasing	0.01, 95% CI [0.00, 0.02]
Direct Behavioral	JCCP	journals	conditional	Clinical	JCCP	0.000	-0.003	0.002	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	JPCS	journals	conditional	Clinical	JPCS	0.000	-0.005	0.003	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	Cognitive	subfields	conditional	Cognitive	hypothetical typical journal	0.000	-0.019	0.014	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Direct Behavioral	Cognition	journals	conditional	Cognitive	Cognition	0.007	0.001	0.022	Increasing	0.01, 95% CI [0.00, 0.02]
Direct Behavioral	JEP:HPP	journals	conditional	Cognitive	JEP:HPP	-0.011	-0.031	-0.002	Decreasing	-0.01, 95% CI [-0.03, -0.00]
Direct Behavioral	JEP:LMC	journals	conditional	Cognitive	JEP:LMC	-0.007	-0.019	-0.001	Decreasing	-0.01, 95% CI [-0.02, -0.00]
Direct Behavioral	JML	journals	conditional	Cognitive	JML	0.000	-0.007	0.006	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	M&C	journals	conditional	Cognitive	M&C	-0.001	-0.011	0.004	No trend detected	-0.00, 95% CI [-0.01, 0.00]
Direct Behavioral	Developmental	subfields	conditional	Developmental	hypothetical typical journal	0.000	-0.025	0.023	No trend detected	-0.00, 95% CI [-0.02, 0.02]
Direct Behavioral	CD	journals	conditional	Developmental	CD	0.005	0.000	0.015	Increasing	0.01, 95% CI [0.00, 0.02]
Direct Behavioral	DP	journals	conditional	Developmental	DP	-0.006	-0.020	0.001	No trend detected	-0.01, 95% CI [-0.02, 0.00]
Direct Behavioral	JECP	journals	conditional	Developmental	JECP	0.000	-0.011	0.012	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	JRA	journals	conditional	Developmental	JRA	0.000	-0.009	0.006	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	PA	journals	conditional	Developmental	PA	0.005	0.000	0.018	No trend detected	0.01, 95% CI [-0.00, 0.02]
Direct Behavioral	Industrial & Organizational	subfields	conditional	Industrial & Organizational	hypothetical typical journal	0.000	-0.020	0.013	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Direct Behavioral	JAP	journals	conditional	Industrial & Organizational	JAP	0.000	-0.004	0.003	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	JEP:A	journals	conditional	Industrial & Organizational	JEP:A	-0.001	-0.010	0.003	No trend detected	-0.00, 95% CI [-0.01, 0.00]
Direct Behavioral	JOB	journals	conditional	Industrial & Organizational	JOB	0.000	-0.005	0.004	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	OBHDP	journals	conditional	Industrial & Organizational	OBHDP	0.002	-0.019	0.016	No trend detected	0.00, 95% CI [-0.02, 0.02]
Direct Behavioral	PP	journals	conditional	Industrial & Organizational	PP	-0.006	-0.028	0.000	Decreasing	-0.01, 95% CI [-0.03, -0.00]
Direct Behavioral	Social & Personality	subfields	conditional	Social & Personality	hypothetical typical journal	0.000	-0.019	0.013	No trend detected	-0.00, 95% CI [-0.02, 0.01]
Direct Behavioral	Emotion	journals	conditional	Social & Personality	Emotion	-0.004	-0.015	0.000	Decreasing	-0.00, 95% CI [-0.02, -0.00]
Direct Behavioral	JESP	journals	conditional	Social & Personality	JESP	0.000	-0.008	0.005	No trend detected	-0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	JPSP	journals	conditional	Social & Personality	JPSP	0.000	-0.007	0.013	No trend detected	0.00, 95% CI [-0.01, 0.01]
Direct Behavioral	JRP	journals	conditional	Social & Personality	JRP	0.000	-0.003	0.002	No trend detected	-0.00, 95% CI [-0.00, 0.00]
Direct Behavioral	PSPB	journals	conditional	Social & Personality	PSPB	-0.004	-0.014	0.004	No trend detected	-0.00, 95% CI [-0.01, 0.00]

Plot field + subfields

# show_col(viridis::mako(25))
# viridis::mako(25)

p_trend_subfields <- posterior_trends |>
  filter(re_level != "journals") |>
  mutate(label = fct_relevel(label,
                             "Psychology (marginal)",
                             "Psychology (conditional)",
                             "General",
                             "Clinical",
                             "Cognitive",
                             "Developmental",
                             "Industrial & Organizational",
                             "Social & Personality")) |>
  ggplot(aes(trend_estimate, outcome, color = trend_category)) +
  geom_vline(xintercept = 0, linetype = "dotted") +
  geom_vline(xintercept = 0, linetype = "dotted") +
  geom_linerangeh(aes(xmin = trend_ci_lower, xmax = trend_ci_upper)) +
  geom_point() +
  #coord_cartesian(xlim = c(-.1, .1)) +
  theme_linedraw() +
  ylab("") +
  xlab("Trend (change per year)") +
  scale_x_continuous(labels = scales::label_percent(), breaks = breaks_pretty()) +
  scale_color_manual(values = c("#38629DFF", "#2B1C35FF", "#359EAAFF")) +
  guides(color = guide_legend(title = "Trend",
                              reverse = TRUE)) +
  facet_wrap(~ label, ncol = 2)

p_trend_subfields

quicksave("p_trend_subfields", 11, 6)

Plot journals

p_trend_journals <- posterior_trends |>
  filter(re_level == "journals") |>
  mutate(label = fct_relevel(label, subfield_and_journal_labels_without_subfields),
         label = fct_rev(label)) |>
  ggplot(aes(trend_estimate, label, color = subfield)) + # color = trend_category, 
  geom_vline(xintercept = 0, linetype = "dotted") +
  geom_linerangeh(aes(xmin = trend_ci_lower, xmax = trend_ci_upper), position = position_dodge(width = .8)) +
  geom_point(position = position_dodge(width = .8)) +
  #coord_cartesian(xlim = c(-.1, .1)) +
  theme_linedraw() +
  ylab("") +
  xlab("Trend (change per year)") +
  scale_x_continuous(labels = scales::label_percent(), breaks = breaks_pretty()) +
  #scale_color_manual(values = c("#38629DFF", "#2B1C35FF", "#359EAAFF")) +
  guides(color = guide_legend(title = "Trend")) +
  facet_wrap(~ outcome, ncol = 3)

p_trend_journals

quicksave("p_trend_journals", 12, 10)

Plot prevalences and trends

Field

# calculate empirical means
data_summary_field <- data_centered |>
  pivot_longer(cols = c("directbehavioral", 
                        "behavioralproxy",
                        "selfreportsaboutbehavior",
                        "selfreportaboutsubjectivestate",
                        "neurobiopsychophys",
                        "cognitiveability"),
               names_to = "outcome",
               values_to = "used") |>
  group_by(year, outcome) |>
  summarize(proportion = mean(used)) |>
  mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",       
                             outcome == "behavioralproxy" ~ "Behavioral Proxy",
                             outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                             outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                             outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                             outcome == "cognitiveability" ~ "Cognitive Ability"),
         outcome = fct_relevel(outcome, 
                               "Direct Behavioral",     
                               "Behavioral Proxy",
                               "Self-Report about Behavior",
                               "Self-Report about Subjective States",
                               "Neuro/Bio/Psychophys",
                               "Cognitive Ability"))

if(exists("models/posterior_prevalence_field_marginal_by_year.rds") &
   exists("models/posterior_prevalence_field_conditional_by_year.rds")){
  
  posterior_prevalence_field_marginal_by_year <- 
    read_rds("models/posterior_prevalence_field_marginal_by_year.rds")
  
  posterior_prevalence_field_conditional_by_year <- 
    read_rds("models/posterior_prevalence_field_conditional_by_year.rds")
  
} else {
  
  # field - marginal effect
  posterior_prevalence_field_marginal_by_year <- 
    expand_grid(data_centered |>
                  distinct(year, year_centered),
                subfield = "across field", 
                journal = "across field",
                type = "marginal",
                re_level = "field") |>
    tibble(res = pmap(list(subfield, journal, type, year_centered), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res) |>
    mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",     
                               outcome == "behavioralproxy" ~ "Behavioral Proxy",
                               outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                               outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                               outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                               outcome == "cognitiveability" ~ "Cognitive Ability"),
           outcome = fct_relevel(outcome, 
                                 "Direct Behavioral",       
                                 "Behavioral Proxy",
                                 "Self-Report about Behavior",
                                 "Self-Report about Subjective States",
                                 "Neuro/Bio/Psychophys",
                                 "Cognitive Ability"))
  
  # field - conditional effect
  posterior_prevalence_field_conditional_by_year <- 
    expand_grid(data_centered |>
                  distinct(year, year_centered),
                subfield = "hypothetical typical subfield", 
                journal = "hypothetical typical journal",
                type = "conditional",
                re_level = "field") |>
    tibble(res = pmap(list(subfield, journal, type, year_centered), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res) |>
    mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",     
                               outcome == "behavioralproxy" ~ "Behavioral Proxy",
                               outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                               outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                               outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                               outcome == "cognitiveability" ~ "Cognitive Ability"),
           outcome = fct_relevel(outcome, 
                                 "Direct Behavioral",       
                                 "Behavioral Proxy",
                                 "Self-Report about Behavior",
                                 "Self-Report about Subjective States",
                                 "Neuro/Bio/Psychophys",
                                 "Cognitive Ability"))
  
  #write_rds(posterior_prevalence_field_year, "models/posterior_prevalence_field_year.rds")
  write_rds(posterior_prevalence_field_marginal_by_year,
            "models/posterior_prevalence_field_marginal_by_year.rds")
  write_rds(posterior_prevalence_field_conditional_by_year, 
            "models/posterior_prevalence_field_conditional_by_year.rds")
}

# plot
p_prevalence_and_trend_field_marginal <- ggplot() +
  geom_ribbon(data = posterior_prevalence_field_marginal_by_year, aes(year, ymin = prevalence_ci_lower, ymax = prevalence_ci_upper), fill = "skyblue", alpha = 0.5) +
  geom_smooth(data = posterior_prevalence_field_marginal_by_year, aes(year, prevalence_estimate),
              method = "lm", color = "black", size = 0.75, se = FALSE) +
  geom_point(data = data_summary_field, aes(year, proportion), size = 1) +
  theme_linedraw() +
  scale_x_continuous(breaks = c(2010, 2012, 2014, 2016, 2018, 2020, 2022)) +
  coord_cartesian(ylim = c(0,1)) +
  scale_y_continuous(labels = scales::label_percent()) + # , breaks = c(0, 0.05, 0.1, 0.25, 0.5, 1)
  ylab("Prevalence\n(Marginal Global Mean)") +
  xlab("Year") +
  facet_wrap( ~ outcome, ncol = 2)

p_prevalence_and_trend_field_marginal

quicksave("p_prevalence_and_trend_field_marginal", 9, 8)


p_prevalence_and_trend_field_conditional <- ggplot() +
  geom_ribbon(data = posterior_prevalence_field_conditional_by_year, aes(year, ymin = prevalence_ci_lower, ymax = prevalence_ci_upper), fill = "skyblue", alpha = 0.5) +
  geom_smooth(data = posterior_prevalence_field_conditional_by_year, aes(year, prevalence_estimate),
              method = "lm", color = "black", size = 0.75, se = FALSE) +
  geom_point(data = data_summary_field, aes(year, proportion), size = 1) +
  theme_linedraw() +
  scale_x_continuous(breaks = c(2010, 2012, 2014, 2016, 2018, 2020, 2022)) +
  coord_cartesian(ylim = c(0,1)) +
  scale_y_continuous(labels = scales::label_percent()) + # , breaks = c(0, 0.05, 0.1, 0.25, 0.5, 1)
  ylab("Prevalence\n(Conditional Global Mean)") +
  xlab("Year") +
  facet_wrap( ~ outcome, ncol = 2)

p_prevalence_and_trend_field_conditional

quicksave("p_prevalence_and_trend_field_conditional", 9, 8)

Subfields

Even more than the other chunks, this one takes time to run. Not run during dev. Change eval=TRUE to run.

# calculate empirical means
data_summary_subfields <- data_centered |>
  pivot_longer(cols = c("directbehavioral", 
                        "behavioralproxy",
                        "selfreportsaboutbehavior",
                        "selfreportaboutsubjectivestate",
                        "neurobiopsychophys",
                        "cognitiveability"),
               names_to = "outcome",
               values_to = "used") |>
  group_by(subfield, year, outcome) |>
  summarize(proportion = mean(used)) |>
  mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",       
                             outcome == "behavioralproxy" ~ "Behavioral Proxy",
                             outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                             outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about\nSubjective Experience",
                             outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                             outcome == "cognitiveability" ~ "Cognitive Ability"),
         outcome = fct_relevel(outcome, 
                               "Direct Behavioral",     
                               "Behavioral Proxy",
                               "Self-Report about Behavior",
                               "Self-Report about\nSubjective Experience",
                               "Neuro/Bio/Psychophys",
                               "Cognitive Ability"))

if(exists("models/posterior_prevalence_subfields_conditional_by_year.rds")){
  
  posterior_prevalence_subfields_conditional_by_year <-
    read_rds("models/posterior_prevalence_subfields_conditional_by_year.rds")
  
} else {
  
  posterior_prevalence_subfields_conditional_by_year <- 
    expand_grid(data_centered |>
                  distinct(year, year_centered, subfield),
                journal = "hypothetical typical journal",
                type = "conditional",
                re_level = "subfields") |>
    tibble(res = pmap(list(subfield, journal, type, year_centered), 
                      est_prevalence, 
                      fit = fitted_model)) |>
    unnest(res)  |>
    mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",     
                               outcome == "behavioralproxy" ~ "Behavioral Proxy",
                               outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                               outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about/nSubjective Experience",
                               outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                               outcome == "cognitiveability" ~ "Cognitive Ability"),
           outcome = fct_relevel(outcome, 
                                 "Direct Behavioral",       
                                 "Behavioral Proxy",
                                 "Self-Report about Behavior",
                                 "Self-Report about\nSubjective Experience",
                                 "Neuro/Bio/Psychophys",
                                 "Cognitive Ability"))
  
  write_rds(posterior_prevalence_subfields_conditional_by_year, 
            "models/posterior_prevalence_subfields_conditional_by_year.rds")
}

# plot
p_prevalence_and_trend_subfields <- ggplot() +
  geom_ribbon(data = posterior_prevalence_subfields_conditional_by_year, aes(year, ymin = prevalence_ci_lower, ymax = prevalence_ci_upper), fill = "skyblue", alpha = 0.5) +
  geom_smooth(data = posterior_prevalence_subfields_conditional_by_year, aes(year, prevalence_estimate),
              method = "lm", color = "black", size = 0.75, se = FALSE) +
  geom_point(data = data_summary_subfields, aes(year, proportion), size = 1) +
  theme_linedraw() +
  #scale_x_continuous(breaks = c(2010, 2012, 2014, 2016, 2018, 2020, 2022)) +
  scale_x_continuous(breaks = c(2010, 2014, 2018, 2022)) +
  coord_cartesian(ylim = c(0,1)) +
  scale_y_continuous(labels = scales::label_percent()) + # , breaks = c(0, 0.05, 0.1, 0.25, 0.5, 1)
  ylab("Prevalence") +
  xlab("Year") +
  facet_grid(subfield ~ outcome)

p_prevalence_and_trend_subfields

quicksave("p_prevalence_and_trend_subfields", 12, 8)

Combined table with summary conclusions

categorical conclusions for field, subfield, and journal levels based on prevalence category and trend (detectably increasing/decreasing, no detectable change)

# join and clean up
posterior_prevalence_and_trends <- 
  full_join(posterior_prevalence, 
            posterior_trends,
            by = c("outcome", "label", "re_level", "type", "subfield", "journal")) |>
  mutate(prevalence_estimate = rnd(prevalence_estimate),
         prevalence_interval = paste0(rnd(prevalence_ci_lower), ", ", rnd(prevalence_ci_upper)),
         trend_estimate = rnd(trend_estimate),
         trend_interval = paste0(rnd(trend_ci_lower), ", ", rnd(trend_ci_upper))) |>
  select(re_level, label, outcome, 
         prevalence_estimate, prevalence_interval, prevalence_category, 
         trend_estimate, trend_interval, trend_category) |>
  arrange(desc(label), desc(outcome))

# save to disk
posterior_prevalence_and_trends |>
  filter(re_level %in% c("field", "subfields")) |>
  select(-re_level) |>
  write_csv("../../data/registered report/results/posterior_prevalence_and_trends_field_subfield.csv")

posterior_prevalence_and_trends |>
  filter(!re_level %in% c("field", "subfields")) |>
  select(-re_level) |>
  write_csv("../../data/registered report/results/posterior_prevalence_and_trends_journals.csv")

# print the table just for field and subfields
posterior_prevalence_and_trends |>
  filter(re_level %in% c("field", "subfields")) |>
  select(-re_level) |>
  kable() |>
  kable_classic(full_width = FALSE)

label	outcome	prevalence_estimate	prevalence_interval	prevalence_category	trend_estimate	trend_interval	trend_category
Psychology (marginal)	Direct Behavioral	0.01	0.01, 0.07	Rare	-0.00	-0.01, 0.00	No trend detected
Psychology (marginal)	Behavioral Proxy	0.05	0.03, 0.10	Uncommon	0.00	-0.01, 0.01	No trend detected
Psychology (marginal)	Self-Report about Behavior	0.07	0.04, 0.15	Uncommon	-0.00	-0.01, 0.01	No trend detected
Psychology (marginal)	Self-Report about Subjective States	0.87	0.68, 0.94	Frequent	0.00	-0.01, 0.02	No trend detected
Psychology (marginal)	Neuro/Bio/Psychophys	0.09	0.05, 0.16	Uncommon	0.00	-0.01, 0.01	No trend detected
Psychology (marginal)	Cognitive Ability	0.05	0.03, 0.08	Rare	0.00	-0.00, 0.01	No trend detected
Psychology (conditional)	Direct Behavioral	0.01	0.00, 0.09	Rare	-0.00	-0.02, 0.01	No trend detected
Psychology (conditional)	Behavioral Proxy	0.05	0.00, 0.30	Rare	0.00	-0.03, 0.04	No trend detected
Psychology (conditional)	Self-Report about Behavior	0.07	0.01, 0.24	Uncommon	-0.00	-0.05, 0.03	No trend detected
Psychology (conditional)	Self-Report about Subjective States	0.89	0.41, 0.99	Frequent	0.00	-0.02, 0.04	No trend detected
Psychology (conditional)	Neuro/Bio/Psychophys	0.10	0.01, 0.34	Uncommon	0.00	-0.03, 0.04	No trend detected
Psychology (conditional)	Cognitive Ability	0.05	0.00, 0.22	Rare	0.00	-0.03, 0.03	No trend detected
General	Direct Behavioral	0.01	0.00, 0.08	Rare	-0.00	-0.02, 0.01	No trend detected
General	Behavioral Proxy	0.05	0.00, 0.30	Rare	-0.00	-0.04, 0.02	No trend detected
General	Self-Report about Behavior	0.04	0.01, 0.16	Rare	-0.00	-0.04, 0.02	No trend detected
General	Self-Report about Subjective States	0.97	0.83, 1.00	Frequent	0.00	-0.01, 0.01	No trend detected
General	Neuro/Bio/Psychophys	0.12	0.01, 0.39	Occasional	0.00	-0.03, 0.04	No trend detected
General	Cognitive Ability	0.04	0.00, 0.22	Rare	0.00	-0.03, 0.02	No trend detected
Clinical	Direct Behavioral	0.01	0.00, 0.08	Rare	-0.00	-0.01, 0.01	No trend detected
Clinical	Behavioral Proxy	0.05	0.00, 0.31	Uncommon	0.01	-0.01, 0.05	No trend detected
Clinical	Self-Report about Behavior	0.06	0.01, 0.20	Uncommon	0.00	-0.04, 0.03	No trend detected
Clinical	Self-Report about Subjective States	0.95	0.71, 0.99	Frequent	0.00	-0.01, 0.02	No trend detected
Clinical	Neuro/Bio/Psychophys	0.09	0.01, 0.33	Uncommon	0.00	-0.03, 0.04	No trend detected
Clinical	Cognitive Ability	0.05	0.00, 0.22	Rare	0.00	-0.03, 0.03	No trend detected
Cognitive	Direct Behavioral	0.01	0.00, 0.09	Rare	-0.00	-0.02, 0.01	No trend detected
Cognitive	Behavioral Proxy	0.05	0.00, 0.32	Uncommon	0.00	-0.02, 0.03	No trend detected
Cognitive	Self-Report about Behavior	0.09	0.01, 0.27	Uncommon	0.00	-0.06, 0.04	No trend detected
Cognitive	Self-Report about Subjective States	0.93	0.62, 0.99	Frequent	-0.00	-0.03, 0.02	No trend detected
Cognitive	Neuro/Bio/Psychophys	0.09	0.01, 0.31	Uncommon	0.00	-0.03, 0.03	No trend detected
Cognitive	Cognitive Ability	0.05	0.00, 0.22	Rare	0.00	-0.03, 0.03	No trend detected
Developmental	Direct Behavioral	0.01	0.00, 0.13	Rare	-0.00	-0.02, 0.02	No trend detected
Developmental	Behavioral Proxy	0.05	0.00, 0.33	Uncommon	-0.00	-0.04, 0.02	No trend detected
Developmental	Self-Report about Behavior	0.09	0.01, 0.27	Uncommon	-0.00	-0.06, 0.03	No trend detected
Developmental	Self-Report about Subjective States	0.78	0.32, 0.96	Frequent	0.01	-0.02, 0.05	No trend detected
Developmental	Neuro/Bio/Psychophys	0.08	0.01, 0.30	Uncommon	0.00	-0.03, 0.03	No trend detected
Developmental	Cognitive Ability	0.05	0.00, 0.23	Rare	0.00	-0.03, 0.03	No trend detected
Industrial & Organizational	Direct Behavioral	0.01	0.00, 0.08	Rare	-0.00	-0.02, 0.01	No trend detected
Industrial & Organizational	Behavioral Proxy	0.05	0.00, 0.29	Rare	0.00	-0.02, 0.03	No trend detected
Industrial & Organizational	Self-Report about Behavior	0.05	0.01, 0.18	Rare	0.00	-0.04, 0.03	No trend detected
Industrial & Organizational	Self-Report about Subjective States	0.78	0.33, 0.97	Frequent	0.00	-0.03, 0.04	No trend detected
Industrial & Organizational	Neuro/Bio/Psychophys	0.10	0.01, 0.36	Occasional	0.00	-0.03, 0.04	No trend detected
Industrial & Organizational	Cognitive Ability	0.05	0.00, 0.23	Rare	0.00	-0.03, 0.03	No trend detected
Social & Personality	Direct Behavioral	0.01	0.00, 0.09	Rare	-0.00	-0.02, 0.01	No trend detected
Social & Personality	Behavioral Proxy	0.04	0.00, 0.29	Rare	0.00	-0.02, 0.03	No trend detected
Social & Personality	Self-Report about Behavior	0.09	0.01, 0.29	Uncommon	-0.00	-0.06, 0.04	No trend detected
Social & Personality	Self-Report about Subjective States	0.80	0.37, 0.97	Frequent	0.01	-0.03, 0.04	No trend detected
Social & Personality	Neuro/Bio/Psychophys	0.09	0.01, 0.32	Uncommon	0.00	-0.03, 0.04	No trend detected
Social & Personality	Cognitive Ability	0.05	0.00, 0.23	Rare	0.00	-0.03, 0.03	No trend detected

Correlations among the random effects

At the level of journals.

Prevalence

# extract correlations/covariances among the random effects
var_cors <- VarCorr(fitted_model)

# wrangle
var_cors_journal <- var_cors$`subfield:journal`$cor |>
  as.data.frame() |>
  rownames_to_column(var = "var")

var_cors_journal_intercept <- var_cors_journal |>
  filter(str_detect(var, "Intercept")) |>
  select(var, contains("Intercept") & contains("Estimate.")) |>
  mutate(var = str_remove(var, "_Intercept")) %>%
  rename_with(~ str_remove(., "Estimate.")) %>%
  rename_with(~ str_remove(., "_Intercept"))

# show_col(viridis::mako(25))
# viridis::mako(25)

# plot
cors_journal_intercept <- var_cors_journal_intercept |>
  rename(outcome = var,
         "Direct Behavioral" = directbehavioral, 
         "Behavioral Proxy" = behavioralproxy, 
         "Self-Report about Behavior" = selfreportsaboutbehavior, 
         "Self-Report about Subjective States" = selfreportaboutsubjectivestate, 
         "Neuro/Bio/Psychophys" = neurobiopsychophys, 
         "Cognitive Ability" = cognitiveability) |>
  mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",       
                             outcome == "behavioralproxy" ~ "Behavioral Proxy",
                             outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                             outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                             outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                             outcome == "cognitiveability" ~ "Cognitive Ability")) 

cors_journal_intercept |>
  mutate_if(is.numeric, rnd) |>
  write_csv("../../data/registered report/results/cors_journal_intercept.csv")

# table
cors_journal_intercept |>
  mutate_if(is.numeric, rnd) |>
  kable(caption = "Correlations among random effects for prevalence at the journal level") |>
  kable_classic(full_width = FALSE)

Correlations among random effects for prevalence at the journal level

outcome	Direct Behavioral	Behavioral Proxy	Self-Report about Behavior	Self-Report about Subjective States	Neuro/Bio/Psychophys	Cognitive Ability
Direct Behavioral	1.00	0.16	0.29	0.21	-0.02	-0.14
Behavioral Proxy	0.16	1.00	0.19	0.05	0.15	-0.31
Self-Report about Behavior	0.29	0.19	1.00	0.29	0.03	-0.06
Self-Report about Subjective States	0.21	0.05	0.29	1.00	0.13	0.03
Neuro/Bio/Psychophys	-0.02	0.15	0.03	0.13	1.00	0.07
Cognitive Ability	-0.14	-0.31	-0.06	0.03	0.07	1.00

p_cors_journal_intercept <- cors_journal_intercept |>
  column_to_rownames("outcome") |>
  ggcorrplot(method = "circle",
             #hc.order = TRUE,
             colors = c("#413E7EFF", "#DEF5E5FF", "#359EAAFF"),
             type = "full",
             outline.col = "grey20",
             ggtheme = ggplot2::theme_linedraw) + 
  ggtitle("Correlations among prevalences")

p_cors_journal_intercept 

quicksave("p_cors_journal_intercept", 7.5, 7.5)

Trend

What was correlation between the change in prevalences over time in the different modes of measurement? As one mode becomes more/less common, what happened for the others?

# wrangle
var_cors_journal_slope <- var_cors_journal |>
  filter(str_detect(var, "_year_centered")) |>
  select(var, contains("_year_centered") & contains("Estimate.")) |>
  mutate(var = str_remove(var, "_year_centered")) %>%
  rename_with(~ str_remove(., "Estimate.")) %>%
  rename_with(~ str_remove(., "_year_centered"))

# show_col(viridis::mako(25))
# viridis::mako(25)

# plot
cors_journal_slope <- var_cors_journal_slope |> 
  rename(outcome = var,
         "Direct Behavioral" = directbehavioral, 
         "Behavioral Proxy" = behavioralproxy, 
         "Self-Report about Behavior" = selfreportsaboutbehavior, 
         "Self-Report about Subjective States" = selfreportaboutsubjectivestate, 
         "Neuro/Bio/Psychophys" = neurobiopsychophys, 
         "Cognitive Ability" = cognitiveability) |>
  mutate(outcome = case_when(outcome == "directbehavioral" ~ "Direct Behavioral",       
                             outcome == "behavioralproxy" ~ "Behavioral Proxy",
                             outcome == "selfreportsaboutbehavior" ~ "Self-Report about Behavior",
                             outcome == "selfreportaboutsubjectivestate" ~ "Self-Report about Subjective States",
                             outcome == "neurobiopsychophys" ~ "Neuro/Bio/Psychophys",
                             outcome == "cognitiveability" ~ "Cognitive Ability")) 

cors_journal_slope |>
  mutate_if(is.numeric, rnd) |>
  write_csv("../../data/registered report/results/cors_journal_slope.csv")

# table
cors_journal_slope |>
  mutate_if(is.numeric, rnd) |>
  kable(caption = "Correlations among random effects for trend at the subfield level") |>
  kable_classic(full_width = FALSE)

Correlations among random effects for trend at the subfield level

outcome	Direct Behavioral	Behavioral Proxy	Self-Report about Behavior	Self-Report about Subjective States	Neuro/Bio/Psychophys	Cognitive Ability
Direct Behavioral	1.00	0.12	-0.09	0.11	0.05	0.48
Behavioral Proxy	0.12	1.00	-0.01	0.00	-0.10	0.13
Self-Report about Behavior	-0.09	-0.01	1.00	-0.20	0.14	-0.21
Self-Report about Subjective States	0.11	0.00	-0.20	1.00	0.13	0.11
Neuro/Bio/Psychophys	0.05	-0.10	0.14	0.13	1.00	0.02
Cognitive Ability	0.48	0.13	-0.21	0.11	0.02	1.00

p_cors_journal_slope <- cors_journal_slope |>
  column_to_rownames("outcome") |>
  ggcorrplot(method = "circle",
             #hc.order = TRUE,
             colors = c("#413E7EFF", "#DEF5E5FF", "#359EAAFF"), 
             type = "full",
             outline.col = "grey20",
             ggtheme = ggplot2::theme_linedraw) + 
  ggtitle("Correlations among trends")

p_cors_journal_slope

quicksave("p_cors_journal_slope", 7.5, 7.5)

Watermark simulated data plots

library(grid)

watermark <- function(fontsize = 20){
  annotation_custom(
    grob = textGrob(label = "Simulated data",
                    gp = gpar(fontsize = fontsize, 
                              col = "red", 
                              alpha = 0.5), 
                    rot = 30),
    xmin = -Inf, xmax = Inf,
    ymin = -Inf, ymax = Inf
  )
}

Subfields

p_prevalence_subfields + watermark(10)

quicksave("p_prevalence_subfields_watermark", 11, 6)

p_trend_subfields + watermark(10)

quicksave("p_trend_subfields_watermark", 11, 6)

Journals

p_prevalence_journals + watermark(30)

quicksave("p_prevalence_journals_watermark", 12, 10)


p_trend_journals + watermark(30)

quicksave("p_trend_journals_watermark", 12, 10)

Prevalences + trends

p_prevalence_and_trend_field_marginal + watermark()

quicksave("p_prevalence_and_trend_field_marginal_watermark", 9, 8)


p_prevalence_and_trend_field_conditional + watermark()

quicksave("p_prevalence_and_trend_field_conditional_watermark", 9, 8)

Even more than the other chunks, this one takes time to run. Not run during dev. Change eval=TRUE to run.

p_prevalence_and_trend_subfields + watermark()

quicksave("p_prevalence_and_trend_subfields_watermark", 12, 8)

Correlations

p_cors_journal_intercept + watermark(30)

quicksave("p_cors_journal_intercept_watermark", 7.5, 7.5)


p_cors_journal_slope + watermark(30)

quicksave("p_cors_journal_slope_watermark", 7.5, 7.5)

Session info

sessionInfo()

## R version 4.3.3 (2024-02-29)
## Platform: aarch64-apple-darwin20 (64-bit)
## Running under: macOS Sonoma 14.5
## 
## Matrix products: default
## BLAS:   /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRblas.0.dylib 
## LAPACK: /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/lib/libRlapack.dylib;  LAPACK version 3.11.0
## 
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
## 
## time zone: Europe/Zurich
## tzcode source: internal
## 
## attached base packages:
## [1] grid      stats     graphics  grDevices utils     datasets  methods  
## [8] base     
## 
## other attached packages:
##  [1] kableExtra_1.4.0       knitr_1.48             marginaleffects_0.14.0
##  [4] emmeans_1.10.2         tidybayes_3.0.6        brms_2.21.0           
##  [7] Rcpp_1.0.12            ggcorrplot_0.1.4.1     ggstance_0.3.7        
## [10] scales_1.3.0           ggplot2_3.5.1          janitor_2.2.0         
## [13] readr_2.1.5            purrr_1.0.2            readxl_1.4.3          
## [16] forcats_1.0.0          stringr_1.5.1          tibble_3.2.1          
## [19] tidyr_1.3.1            dplyr_1.1.4           
## 
## loaded via a namespace (and not attached):
##  [1] gridExtra_2.3        inline_0.3.19        sandwich_3.1-0      
##  [4] rlang_1.1.4          magrittr_2.0.3       multcomp_1.4-25     
##  [7] snakecase_0.11.1     matrixStats_1.0.0    compiler_4.3.3      
## [10] mgcv_1.9-1           loo_2.6.0            reshape2_1.4.4      
## [13] systemfonts_1.0.6    vctrs_0.6.5          crayon_1.5.2        
## [16] pkgconfig_2.0.3      arrayhelpers_1.1-0   fastmap_1.1.1       
## [19] backports_1.4.1      labeling_0.4.3       utf8_1.2.4          
## [22] rmarkdown_2.26       tzdb_0.4.0           ragg_1.3.0          
## [25] bit_4.0.5            xfun_0.46            cachem_1.0.8        
## [28] jsonlite_1.8.8       collapse_2.0.13      highr_0.11          
## [31] parallel_4.3.3       R6_2.5.1             bslib_0.7.0         
## [34] stringi_1.8.4        StanHeaders_2.32.6   lubridate_1.9.3     
## [37] jquerylib_0.1.4      cellranger_1.1.0     estimability_1.5.1  
## [40] rstan_2.32.6         zoo_1.8-12           bayesplot_1.10.0    
## [43] Matrix_1.6-5         splines_4.3.3        timechange_0.3.0    
## [46] tidyselect_1.2.1     rstudioapi_0.16.0    abind_1.4-5         
## [49] yaml_2.3.8           codetools_0.2-20     curl_5.2.1          
## [52] pkgbuild_1.4.4       plyr_1.8.9           lattice_0.22-6      
## [55] withr_3.0.1          bridgesampling_1.1-2 posterior_1.4.1     
## [58] coda_0.19-4.1        evaluate_0.23        survival_3.5-8      
## [61] RcppParallel_5.1.7   xml2_1.3.6           ggdist_3.3.2        
## [64] pillar_1.9.0         tensorA_0.36.2       checkmate_2.3.1     
## [67] stats4_4.3.3         insight_0.19.10      distributional_0.4.0
## [70] generics_0.1.3       vroom_1.6.5          hms_1.1.3           
## [73] rstantools_2.4.0     munsell_0.5.1        xtable_1.8-4        
## [76] glue_1.7.0           tools_4.3.3          data.table_1.15.4   
## [79] mvtnorm_1.2-5        QuickJSR_1.1.3       colorspace_2.1-1    
## [82] nlme_3.1-164         cli_3.6.3            textshaping_0.3.7   
## [85] fansi_1.0.6          svUnit_1.0.6         viridisLite_0.4.2   
## [88] svglite_2.1.3        Brobdingnag_1.2-9    V8_4.3.3            
## [91] gtable_0.3.5         sass_0.4.9           digest_0.6.35       
## [94] TH.data_1.1-2        farver_2.1.2         htmltools_0.5.8.1   
## [97] lifecycle_1.0.4      bit64_4.0.5          MASS_7.3-60.0.1